Continuum Companion to Discourse Analysis 9781441106957, 1441106952

Table of contents :
Cover
HalfTitle
Series
Title
Copyright
Contents
Notes on Contributors
Acknowledgements
Introduction
Part I Methods of Analysis in Discourse Research
1 Data Collection and Transcription in Discourse Analysis
Data Collection as Mediated Action
Five Processes of Entextualization
Data in the Audio Age
Video Killed the Discourse Analyst?
Data Collection and Transcription in the Digital Age
Conclusion
Key Readings
2 Conversation Analysis
Introduction: Origins and Key Features
Data Analysis: The ‘Building Blocks’ of Social Life
The Turn-taking System
Collaborative Completion
Summary, Conclusion and Future Prospects
Transcription Key
Note
Key Readings
3 Critical Discourse Analysis
Introduction
Discourse, Power, Ideology and Critique
Current Approaches and Developments in CDA
Sample Study: Inclusion and Exclusion – The Austrian Case
Summary
Notes
Key Readings
4 Genre Analysis
Current Theory and Research in Genre Analysis
Persuasion in the High-stakes World of Grant Funding: A Sample Genre Analysis
Future Directions of Genre Analysis
Note
Key Readings
5 Narrative Analysis
Introduction
DA Approaches to Narrative
CA Approaches to Narrative
LE Approaches to Narrative
Current Issues and Approaches in Narrative Analysis
Case One: Narrative, Interaction and Identity in a Friendship Group
Case Two: Stance, Positioning and Alignment in Narratives of Professional Experience
Conclusion and Implications
Transcription Key
Key Readings
6 Ethnography and Discourse Analysis
Locating Ethnography
Locating Discourse Analysis
Discourse Analysis and Ethnography: Complementary or in Opposition?
Three Approaches to Discourse Analysis and Ethnography
Discourse Analysis in Critical Ethnography: The Tradewinds Study
Conclusion
Transcription Key
Notes
Key Readings
7 Systemic Functional Linguistics
Synthesis of Current Thinking and Research
Sample Study
New Directions
Notes
Key Readings
8 Multimodal Discourse Analysis
Introduction
Approaches to MDA
Theoretical and Analytical Issues in MDA
Sample MDA Text Analysis
New Directions in MDA
Notes
Key Readings
9 Corpus Approaches to the Study of Discourse
Introduction
Researching Language in Use through Corpus Analysis
A Sample Study: Language in Use
Investigating Language Structure beyond the Sentence and Future Directions for Corpus-based Discourse Analysis
Note
Key Readings
Part II Research Areas and New Directions in Discourse Research
10 Spoken Discourse
Types of Spoken Discourse
Spoken Discourse Features
Function of Features
Social Variables
Sample Study
Conclusion
Transcription Key
Key Readings
11 Academic Discourse
What Is Academic Discourse?
Why Is Academic Discourse Important?
How Is Academic Discourse Studied?
What Do We Know about Academic Discourse?
A Sample Study: Citations
Conclusion
Key Readings
12 Discourse and the Workplace
Introduction
Current Research on Spoken Workplace Discourse
A Sample Study
New Directions in Workplace Discourse Research
Transcription Key
Notes
Key Readings
13 Discourse and Gender
Introduction
The Turn to Discourse
Discourse-based Approaches in Gender and Language Research
A Sample Study: Cougars
Future Directions
Notes
Key Readings
14 Discourse and the News
Introduction
Defining News and News Discourse
Critical Approaches to the Discourse of Newspapers: Ideologies at Work
Discourse and Power: The Broadcast News Interview
Beyond Considerations of Power and Ideology in News Discourse?
Note
Key Readings
15 Discourse and Computer-mediated Communication
Introduction
The Online Context
The New Literacy Studies
The Fabric of the Text: Multimodality and Semiotic Resources
Conclusions and Future Directions
Notes
Key Readings
16 Forensic Discourse Analysis: A Work in Progress
Origins of Forensic Discourse Analysis
The Focus of Forensic Discourse Analysis
The Location of Language within Sociolinguistic Interaction
Observations on the Structuring of Forensic Discourse
The Construction of Admission
Textual Provenance and Mode
Western Australian Royal Commission into Police Corruption
Contemporary Interrogation Methods
Consensus in Court
Summary
Notes
Key Readings
17 Discourse and Identity
Introduction
Defining ‘Minority’ Identity
A Sample Study
Key Readings
18 Discourse and Race
Introduction
Discursive Construction of Race: Production of Self and Other
Whiteness Studies and its Critiques
Elite Discourse and Racism
Everyday Discourse and the New Racism
Positioning Theory and Storyline Analysis of Racist Online Discourses (Lin and Tong 2009)
Conclusion
Note
Key Readings
19 Classroom Discourse
Introduction
Current Thinking and Research on Classroom Discourse
Analyses of ‘turns, sequences and meanings’
A Sample Study
Future Directions
Key Readings
20 Discourse and Intercultural Communication
What Is Culture?
Intercultural Communication
Researching Cultural Values
Intercultural Communicative Competence
A Sample Study: Managing ‘Rapport’ across Cultures
Conclusions
Key Readings
21 Medical Discourse
Introduction
Illness-ing
Patient-ing and Doctor-ing: The Social Epistemics of Knowledge and Sensations
Doing: Being a ‘Competent Patient’
Widening Out the Realm of Medical Discourse Research
Conclusion
Notes
Key Readings
Glossary
References
Author Index
Subject Index

Citation preview

Bloomsbury Companion to Discourse Analysis

HylandK_Prelims_Finals.indd i

10/30/2012 10:17:18 AM

Bloomsbury Companions Continuum Companion to Historical Linguistics, Edited by Silvia Luraghi and Vit Bubenik Available in Paperback as Bloomsbury Companion to Historical Linguistics Continuum Companion to Phonology, Edited by Nancy C. Kula, Bert Botma and Kuniya Nasukawa Available in Paperback as Bloomsbury Companion to Phonology Continuum Companion to Second Language Acquisition, Edited by Ernesto Macaro Available in Paperback as Bloomsbury Companion to Second Language Acquisition Continuum Companion to Discourse Analysis, Edited by Ken Hyland and Brian Paltridge Available in Paperback as Bloomsbury Companion to Discourse Analysis Continuum Companion to the Philosophy of Language, Edited by Manuel García-Carpintero and Max Kölbel Available in Paperback as Bloomsbury Companion to Philosophy of Language Bloomsbury Companion to Syntax, Edited by Edited by Silvia Luraghi and Claudia Parodi

Forthcoming: Bloomsbury Companion to Phonetics, Edited by Rachael-Anne Knight and Mark J. Jones Bloomsbury Companion to TESOL, Edited by Jun Liu Bloomsbury Companion to Lexicography, Edited by Howard Jackson Bloomsbury Companion to M. A. K. Halliday, Edited by Jonathan J. Webster Bloomsbury Companion to Cognitive Linguistics, Edited by John R. Taylor and Jeanette Littlemore Bloomsbury Companion to Stylistics, Edited by Violeta Sotirova Bloomsbury Companion to Translation Studies, Edited by John Kearns and Jody Bryne

HylandK_Prelims_Finals.indd ii

10/30/2012 10:17:18 AM

Bloomsbury Companion to Discourse Analysis Edited by

Ken Hyland and Brian Paltridge

Bloomsbury Companions

L ON DON • N E W DE L H I • N E W Y OR K • SY DN EY

HylandK_Prelims_Finals.indd iii

10/30/2012 10:17:18 AM

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc 50 Bedford Square London WC1B 3DP UK

175 Fi h Avenue New York NY 10010 USA

www.bloomsbury.com First published 2011 by Continuum © Ken Hyland, Brian Paltridge and Contributors, 2011, 2013 All rights reserved. No part of this publication may be reproduced or transmi ed in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. No responsibility for loss caused to any individual or organization acting on or refraining from action as a result of the material in this publication can be accepted by Bloomsbury Academic or the author. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. EISBN: 9781441106957 Library of Congress Cataloging-in-Publication Data A catalogue record for this book is available from the Library of Congress

Typeset by Newgen Imaging Systems Pvt Ltd, Chennai, India

HylandK_Prelims_Finals.indd iv

10/30/2012 10:17:18 AM

Contents

Notes on Contributors Acknowledgements Introduction Ken Hyland and Brian Paltridge

vii xv 1

Part I Methods of Analysis in Discourse Research 1 Data Collection and Transcription in Discourse Analysis Rodney H. Jones

9

2 Conversation Analysis Sue Wilkinson and Celia Kitzinger

22

3 Critical Discourse Analysis Ruth Wodak

38

4 Genre Analysis Christine M. Tardy

54

5 Narrative Analysis Mike Baynham

69

6 Ethnography and Discourse Analysis Dwight Atkinson, Hanako Okada and Steven Talmy

85

7 Systemic Functional Linguistics J. R. Martin

101

8 Multimodal Discourse Analysis Kay L. O’Halloran

120

9 Corpus Approaches to the Study of Discourse Bethany Gray and Douglas Biber

138

HylandK_Prelims_Finals.indd v

12/1/2010 8:11:25 AM

Contents

Part II Research Areas and New Directions in Discourse Research 10 Spoken Discourse Joan Cutting

155

11 Academic Discourse Ken Hyland

171

12 Discourse and the Workplace Janet Holmes

185

13 Discourse and Gender Paul Baker

199

14 Discourse and the News Martin Montgomery

213

15 Discourse and Computer-mediated Communication Julia Davies

228

16 Forensic Discourse Analysis: A Work in Progress John Olsson

244

17 Discourse and Identity Tope Omoniyi

260

18 Discourse and Race Angel Lin and Ryuko Kubota

277

19 Classroom Discourse Jennifer Hammond

291

20 Discourse and Intercultural Communication John Corbett

306

21 Medical Discourse Timothy Halkowski

321

Glossary References Author Index Subject Index

333 337 375 383

vi

HylandK_Prelims_Finals.indd vi

12/1/2010 8:11:25 AM

Notes on Contributors

Dwight Atkinson is an applied linguist and second language educator who specializes in writing, qualitative research approaches to the study of language, and second language acquisition Current projects include developing a ‘sociocognitive’ approach to second language acquisition and research on the experiences of vernacular language-schooled students in English-language universities in India. Atkinson’s past work has ranged widely, from the history of medical and scientific research writing in English, to critiques of concepts in writing instruction such as critical thinking and voice, to explorations of the concept of culture, to writings on qualitative research methods. Atkinson teaches courses in qualitative research and second language acquisition at Purdue University. Paul Baker is a Senior Lecturer in the Department of Linguistics and English Language at Lancaster University. His books include Sociolinguistics and Corpus Linguistics (2010), Sexed Texts: Language, Gender and Sexuality (2008), Using Corpora in Discourse Analysis (2006) and Public Discourses of Gay Men (2005). He is the commissioning editor for the journal Corpora. He has published journal articles examining the representation of identities that include Muslims, refugees and asylum seekers and gay men. He is particularly interested in the use of corpus methods in order to examine identities. Mike Baynham is Professor of TESOL at the University of Leeds. A sociolinguist by training and applied linguist by affiliation, he has had since his doctoral research a long standing interest in narrative, particularly migration narratives, and has published a number of papers on narrative topics. He is co-convenor of the AILA Research Network on Language and Migration. In 2005 he edited with Anna de Fina the collection Dislocations/Relocations: Narratives of Displacement. Douglas Biber is Regents’ Professor of English (Applied Linguistics) at Northern Arizona University. His research efforts have focused on corpus linguistics, English grammar, and register variation (in English and across languages)

HylandK_Prelims_Finals.indd vii

12/1/2010 8:11:25 AM

Notes on Contributors

as well as in synchronic and diachronic change. He has written numerous books and monographs, including academic books published with Cambridge University Press (1988, 1995, 1998, 2009), John Benjamins (2006, 2007), and the co-authored Longman Grammar of Spoken and Written English (1999). John Corbett is Professor of English at the University of Macau. He is the author of various articles and books on language education, including An Intercultural Approach to English Language Teaching and, with Wendy Anderson, Exploring English with Online Corpora. He directs the Scottish Corpus of Texts and Speech project at www.scottishcorpus.ac.uk. Joan Cutting is a Senior Lecturer in TESOL in the University of Edinburgh, specializing in EFL methodology, TESOL materials design, and text and discourse for TESOL. She researches spoken English, her main interests being vague language and in-group code, with a special emphasis on cross-cultural differences, and also international students’ interactions in spoken domains in UK Higher Education. She considers the implications for EFL/EAL/EAP teachers and for non-language-specialist staff development. She is the author of Analysing the Language of Discourse Communities (2000), Vague Language Explored (2007) and Pragmatics and Discourse (2008), and co-editor of the Edinburgh Textbooks in TESOL series (2011–2015). Julia Davies is based in the School of Education at the University of Sheffield in the United Kingdom where she directs the Literacy and Language EdD and the MA in New Literacies. Her research explores online digital interactivity and text production, focusing on how language and literacy, our sense of our selves and the world we live in are being affected by these practices. Recent work includes a book (with Guy Merchant) Web 2.0 for Schools: Learning and Social Participation (Peter Lang). Bethany Gray is a PhD candidate in Applied Linguistics at Northern Arizona University. Her research investigates register variation using corpus linguistics methodologies, with a focus on documenting variation across written academic registers and disciplines. Her research has also included investigations into stance, shell nouns, historical change in research writing, and spoken and written register variation. These research efforts have resulted in published book chapters and research articles in journals such as Journal of English for Academic Purposes, English for Specific Purposes and TESOL Quarterly. Timothy Halkowski is a professor in the Division of Communication at the University of Wisconsin – Stevens Point. His research has focused on interactional uses of the concept of ‘role’, ways in which answers to questions are viii

HylandK_Prelims_Finals.indd viii

12/1/2010 8:11:25 AM

Notes on Contributors

interactionally generated, the interactional accomplishment of coherence in aphasic discourse, and how patients convey themselves to be reliable and reasonable reporters of their physical symptoms. Currently he is researching how the quantification of tobacco and alcohol use is managed in doctor–patient discourse. He also regularly serves as a grant reviewer for the National Institutes of Health in the United States. Jennifer Hammond is an Associate Professor and Honorary Associate in the Faculty of Arts and Social Sciences (Education), University of Technology, Sydney. She has taught for many years in the fields of language and literacy education, ESL education and research design. Her research interests are in literacy development; classroom interaction, and the implications of sociocultural and systemic theories of language and learning in ESL education. She has published widely in these areas. Janet Holmes holds a personal Chair in Linguistics at Victoria University of Wellington. She teaches sociolinguistics courses, specializing in workplace discourse, New Zealand English, and language and gender. She is Director of the Wellington Language in the Workplace project (see www.victoria.ac.nz/lals/ lwp) and a Fellow of the Royal Society of New Zealand. Her books include An Introduction to Sociolinguistics, now in its third edition, the Blackwell Handbook of Language and Gender (with Miriam Meyerhoff) and Gendered Talk at Work. She has also published articles in international journals on topics ranging from gendered discourse, through socio-pragmatic aspects of interaction, to variationist features of New Zealand English. Ken Hyland is Chair Professor of Applied Linguistics and Director of the Centre for Applied English Studies at the University of Hong Kong. He has taught Applied Linguistics and EAP for over 30 years in Asia, Australasia and, most recently, at the Institute of Education in London, and has published over 140 articles and 14 books on language education and academic writing. He was founding co-editor of the Journal of English for Academic Purposes and is now coeditor of Applied Linguistics and editor of the Continuum Discourse Series. Rodney H. Jones is Associate Professor of English at City University of Hong Kong. His research interests include mediated discourse analysis, qualitative research methods and language and sexuality. He is co-editor of Discourse in Action: Introducing Mediated Discourse Analysis (Routledge) and Advances in Discourse Studies (Routledge) and editor of the soon to be published volume Discourse and Creativity (Longman). Celia Kitzinger is Professor of Gender, Sexuality and Conversation Analysis in the Department of Sociology at the University of York, United Kingdom. ix

HylandK_Prelims_Finals.indd ix

12/1/2010 8:11:25 AM

Notes on Contributors

She has published nine books and around 200 articles and book chapters on issues related to gender, sexuality and language. Her current research interests include the analysis of interactions on childbirth helplines and (with Gene Lerner) repair practices in ordinary interaction. Ryuko Kubota is a professor in the Department of Language and Literacy Education in the Faculty of Education at the University of British Columbia, Vancouver, Canada. Her main area of research is critical applied linguistics with a focus on culture, multiculturalism, written discourse, race, and critical pedagogy. Her articles have appeared in such journals as Canadian Modern Language Review, Critical Inquiry in Language Studies, Foreign Language Annals, Journal of Second Language Writing, Modern Language Journal, TESOL Quarterly, Written Communication and World Englishes. She is an editor of Race, Culture, and Identities in Second Language: Exploring Critically Engaged Practice (Routledge, 2009). Angel Lin is an Associate Professor in the Faculty of Education, the University of Hong Kong. She works in the areas of language and identity studies, bilingual education, new literacies and youth cultural studies. She has published six research books and over 70 book chapters and research articles. She serves on the editorial boards of a number of international research journals in the field including: Applied Linguistics, Journal of Critical Discourse Studies, Pragmatics and Society, International Journal of Bilingual Education and Bilingualism, and is an Associate Editor of Linguistics and Education. J. R. Martin is Professor of Linguistics at the University of Sydney. His research interests include systemic theory, functional grammar, discourse semantics, register, genre, multimodality and critical discourse analysis, focusing on English and Tagalog. Recent publications include The Language of Evaluation (with Peter White) (Palgrave, 2005); with David Rose, a second edition of Working with Discourse (Continuum, 2007) and a book on genre, Genre Relations: Mapping Culture (Equinox, 2008); a second edition of the 1997 Functional Grammar Workbook; and with Clare Painter and Christian Matthiessen, Deploying Functional Grammar (Commercial Press, Beijing, 2010). The first four volumes of his collected papers (edited by Wang Zhenhua, Shanghai Jiaotong University Press) are being published in 2010. Martin Montgomery has published widely on sociolinguistics, stylistics and discourse analysis. He is the author of The Discourse of Broadcast News (Routledge, 2007) and of An Introduction to Language and Society (3rd edition, Routledge, 2008). After serving successively as Director of the Programme in Literary Linguistics, Head of the Department of English Studies and Director x

HylandK_Prelims_Finals.indd x

12/1/2010 8:11:26 AM

Notes on Contributors

of the Scottish Centre of Journalism Studies at the University of Strathclyde, Glasgow, he has recently been appointed Chair Professor in English Linguistics and head of the Department of English at the University of Macau, China. Kay L. O’Halloran is Director of the Multimodal Analysis Lab in the Interactive and Digital Media Institute (IDMI) and Associate Professor in the Department of English Language and Literature at the National University of Singapore. Her research fields include multimodal discourse analysis, with a specific interest in multimodal approaches to mathematics and the development of interactive digital technology for multimodal analysis. She is Principal Investigator for several large projects in the Multimodal Analysis Lab IDMI (http://multimodal-analysis-lab.org/). Hanako Okada teaches academic writing to students from diverse linguistic and cultural backgrounds at Sophia University in Tokyo, Japan. Her research interests are in situated qualitative research, reflective personal narratives, bilingual identity and sociocognitive approaches to second language acquisition. Past research has been published in the Modern Language Journal and Learning the Literacy Practices of Graduate School: Insiders’ Reflections on Academic Enculturation. Current research projects include narratives of the chronically ill, ‘pleasures and perils’ of narrative inquiry, and the issue of labels in applied linguistics and TESOL. John Olsson is the founder of the Forensic Linguistics Institute, the world’s first dedicated forensic linguistics laboratory for the analysis of language which occurs in criminal and civil allegations and trials. Since 1994 he has conducted over 350 investigations into authorship, the analysis of discourse and questioned utterances. These investigations have related to all major crimes including homicide, terrorism, kidnap, blackmail, organized crime, assault of all types, witness intimidation, and people and narcotic trafficking. He has appeared many times as an expert witness in a wide variety of courts throughout the English speaking world. He is the author of Forensic Linguistics, Wordcrime, and a large number of book chapters and journal articles. He is the Coordinator of the Forensic Linguistics programme at Bangor University, Wales. He holds postgraduate degrees in Linguistics from Bangor, Birmingham and Glamorgan universities. Tope Omoniyi is Professor of Sociolinguistics at Roehampton University, London. His area of research interest is language and identity which he has explored in the context of language education, borderlands, popular culture and religion. He is Director of the Centre for Research in English Language and Linguistics and coordinator of the MPhil/PhD programmes. His publications xi

HylandK_Prelims_Finals.indd xi

12/1/2010 8:11:26 AM

Notes on Contributors

include The Sociolinguistics of Borderlands: Two Nations, One Community (Africa World Press, 2004), and the co-edited volumes Explorations in the Sociology of Language and Religion (John Benjamins, 2006), The Sociolinguistics of Identity (Continuum, 2006) and Contending with Globalization in World Englishes (Multilingual Matters, 2010). Brian Paltridge is Professor of TESOL at the University of Sydney. He is author of Genre, Frames and Writing in Research Settings (John Benjamins, 1997), Making Sense of Discourse Analysis (AEE Publishers, 2000), Genre and the Language Learning Classroom (University of Michigan Press, 2001), Discourse Analysis (Continuum, 2006), with Sue Starfield Thesis and Dissertation Writing in a Second Language (Routledge, 2007), and with his TESOL colleagues at the University of Sydney, Teaching Academic Writing (University of Michigan Press, 2009). With Diane Belcher and Ann Johns he edited New Directions for ESP Research (University of Michigan Press, 2011) and, with Sue Starfield, The Handbook of English for Specific Purposes for the Blackwell Handbooks in Linguistics series. Steven Talmy is an Assistant Professor in the Department of Language and Literacy Education at the University of British Columbia. His research on high school students’ resistance to ESL, and negative language ideologies concerning ESL, has appeared in such journals as Applied Linguistics, Linguistics and Education, and Journal of Language, Identity, and Education, as well as several edited anthologies. His academic interests include critical analyses of discourse, K-12 ESL, the sociology of ESL education, and qualitative research methods, especially research interviews. Christine M. Tardy is an Associate Professor of Writing, Rhetoric and Discourse at DePaul University in Chicago. Her work on genre and second language writing appears in English for Specific Purposes, Journal of Second Language Writing, Journal of English for Academic Purposes, Discourse and Society and Written Communication. Her book Building Genre Knowledge (Parlor Press, 2009) develops a methodological and theoretical framework for the study of genre learning over time. Sue Wilkinson is Professor of Feminist and Health Studies in the Social Sciences Department at Loughborough University, United Kingdom. She is the founding editor of the international journal, Feminism & Psychology and her publications encompass six books and more than a hundred articles in the areas of gender/sexuality, feminism, health and qualitative methods. She has a longstanding interest in the social construction of inequality and is a campaigner for equal marriage rights for same-sex couples. She (re)trained in conversation analysis (CA) at UCLA in 2001–2002 and her recent work uses CA xii

HylandK_Prelims_Finals.indd xii

12/1/2010 8:11:26 AM

Notes on Contributors

to study helpline interaction. She is particularly interested in technical specifications of repair. Ruth Wodak is Distinguished Professor of Discourse Studies at Lancaster University. Her research interests focus on discourse studies; gender studies; language and/in politics; prejudice and discrimination; and on ethnographic methods of linguistic fieldwork. She is member of the editorial board of a range of linguistic journals and co-editor of the journals and co-editor of the book series Discourse Approaches to Politics and Society and Culture. Recent book publications include Qualitative Discourse Analysis in the Social Sciences (with M. Krzyżanowski, 2008); Migration, Identity and Belonging (with G. Delanty, P. Jones, 2008), The Discursive Construction of History. Remembering the Wehrmacht’s War of Annihilation (with H. Heer, W. Manoschek, A. Pollak, 2008), The Politics of Exclusion (with M. Krzyżanowski, 2009), Gedenken im Gedankenjahr (with R. de Cillia, 2009) and The Discourse of Politics in Action: ‘Politics as Usual’ (2009).

xiii

HylandK_Prelims_Finals.indd xiii

12/1/2010 8:11:26 AM

HylandK_Prelims_Finals.indd xiv

12/1/2010 8:11:26 AM

Acknowledgements

The editors would like to thank Gurdeep Mattu of Continuum for proposing this project and for his support throughout its development. They would also like to thank Mr P. Muralidharan, head of the project management team at Newgen Imaging Systems of Chennai, India, for the very careful work in the copyediting, typesetting, proofreading and layout stages of the book. Finally, we want to thank the authors for agreeing to write the chapters and for seeing the process through to a successful conclusion. We really are delighted with them and hope they enjoyed the process as much as we did.

HylandK_Prelims_Finals.indd xv

12/1/2010 8:11:26 AM

HylandK_Prelims_Finals.indd xvi

12/1/2010 8:11:26 AM

Introduction Ken Hyland and Brian Paltridge

Chapter Overview Methods and Approaches Areas of Research Further Reading

2 3 4

Discourse is one of the most significant concepts of modern thinking in a range of disciplines across the humanities and social sciences. This is because it concerns the ways that language works in our engagements with the world and our interactions with each other, so creating and shaping the social, political and cultural formations of our societies. From an applied linguistic perspective, to study discourse is therefore to study language in action, looking at texts in relation to the social contexts in which they are used. But because language is connected to almost everything that goes on in the world, ‘discourse’ is something of an overloaded term, covering a range of meanings. People who study discourse might therefore focus on the analysis of speech and writing to bring out the dynamics and conventions of social situations, or take a more theoretical and critical point of view to consider the institutionalized ways of thinking that define our social lives. Discourse, in fact, can be seen to spread between two poles, giving moreor-less emphasis to concrete texts or to institutional practices, to either particular cases of talk or to how social structures are formed by it. In this book we seek to provide a way into this complex and wide-ranging field for beginning researchers in the area of applied linguistics. Through a collection of chapters specifically commissioned by experts in different areas of the field, we attempt to offer an accessible and authoritative introduction to the many facets of this fascinating and complex topic. Together, these chapters provide teachers, students and researchers with a way of theorizing and investigating both spoken and written discourse. The book is in two parts. The first is on methods of analysis and contains chapters which explore and describe the main approaches and issues in 1

HylandK_Intro_Finals.indd 1

12/1/2010 8:10:57 AM

Bloomsbury Companion to Discourse Analysis

researching discourse. Topics that the authors have addressed include assumptions which underlie the particular method or approach, issues of validity or trustworthiness in research techniques and instruments appropriate to the goal and method of research. The second part provides an overview and discussion of a key area of discourse studies, addressing the main methods of investigation and central issues and findings. In each chapter the authors have included a sample study that illustrates the points they are making in their discussion. The authors have also identified resources for further reading on the particular method or issue under discussion.

Methods and Approaches In the first chapter in the volume, Rodney Jones discusses issues in the collection and transcription of spoken data. As he points out, data collection and the approach to transcription are affected by the theoretical interests of the analyst. These will influence the texts that are chosen for analysis and the aspects of the interaction/s that will be examined. This is an extremely important point as it applies to each of the methods of analysis and research areas covered in the book. Sue Wilkinson and Celia Kitzinger, in their chapter on conversation analysis, provide many examples of the fine-grained analysis of spoken language and what it can reveal about how social relations are developed through the use of spoken discourse. The chapter which follows, by Ruth Wodak, discusses how critical discourse analysis can reveal norms and values and relations of power which underlie texts. Through a discussion of images that have been obtained from the internet she shows how issues such as racism, discrimination and exclusion are produced and reproduced through discourse and the use of visual metaphors and symbols. Christine Tardy’s chapter on genre analysis draws on work in English for specific purposes and rhetorical studies for the background to her analysis. She outlines a range of genre analysis methods, and then presents a study of the project summary component of grant funding applications to illustrate the value of this approach to the examination of texts. The chapter by Mike Baynham on narrative analysis describes an approach to discourse analysis that has a history which dates back to work in folklore studies, semiotics and the study of myth. He examines three approaches to narrative analysis: discourse analytic approaches, conversation analysis and linguistic ethnographic approaches to the analysis of narrative. The chapter on ethnography and discourse analysis by Dwight Atkinson, Hanako Okada and Steven Talmy discusses three approaches to an ethnographic analysis of discourse: the ethnography of communication, microethnography and critical ethnography. The sample study they choose to illustrate the points they make examines English 2

HylandK_Intro_Finals.indd 2

12/1/2010 8:10:58 AM

Introduction

as a Second Language as a stigmatized identity category at a large public high school in Hawai’i. Jim Martin’s chapter on systemic functional linguistics provides a detailed account of the theoretical background to this approach to discourse analysis. Through the examination of a newspaper text, Martin provides an illustration of how this view of language and social context can be applied to the analysis of everyday texts and what it can reveal about their construction. Also working within the systemic functional tradition, Kay L. O’Halloran, in her chapter on multimodal discourse analysis, presents an analysis of a television programme in which modalities other than just language play an important part in the way the text progresses. The chapter on corpus studies by Bethany Gray and Douglas Biber describes how corpus linguistics can be used to analyse discourse. Through an analysis of conversation and academic writing, they show how corpus analysis can provide insights into how language varieties are constructed and realized linguistically.

Areas of Research The first chapter in this section, by Joan Cutting, examines the analysis of spoken discourse. Her chapter describes types of spoken discourse and social variables that influence the choice of spoken discourse features. The study she chooses to illustrate the points she makes is an analysis of the spoken interaction of international students in UK universities. In his chapter on academic discourse, Ken Hyland discusses reasons for studying academic discourse as well as providing a detailed account of how academic discourse is studied. He looks at disciplinary differences in citation practices in order to give an example of analysis in the area and what it can tell us about approaches to academic persuasion. In the chapter on discourse and the workplace, Janet Holmes provides an overview of current research on spoken workplace discourse. She then describes a study which is part of the Wellington Language in the Workplace Project to illustrate the points she discusses. The chapter on discourse and gender by Paul Baker offers a history of developments in this area of research, then examines the notion of ‘cougars’ as an emerging gendered identity. Using a large-scale corpus of texts including newspaper articles, transcripts of television chat shows and fiction, he looks for examples of people talking about ‘cougars’ and what this term refers to in the texts. In his chapter on discourse and the news, Martin Montgomery discusses studies that examine news from both structural and ideological points of view. He illustrates this by an analysis of ways of expressing blame in newspaper reports and ways in which newspaper articles use the expressions war on terror and war on terrorism. The chapter on discourse and computer mediated 3

HylandK_Intro_Finals.indd 3

12/1/2010 8:10:58 AM

Bloomsbury Companion to Discourse Analysis

communication by Julia Davies discusses blogs and textual presentations of the self in these online texts. She analyses in particular, how individuals present themselves in blogs and connect with others through their use of discourse. John Olsson’s chapter on discourse and forensics discusses the origins of forensic discourse analysis, its location within the sociolinguistic tradition of discourse analysis as well as ways in which the forensic analysis of discourse has been drawn on in legal proceedings. Tope Omoniyi’s chapter on discourse and identity reviews research in this area before describing a project which examines issues of race and identity in minority newspapers in the United Kingdom. Omoniyi shows how minority identities are often placed in contrast to the powerful mainstream in media discourse and how the essence of minority identity derives from the existence of an identifiable and privileged majority. Discourse and race is discussed by Angel Lin and Ryuko Kubota who draw on insights from cultural studies, critical theory, postcolonial studies and discourse analysis to examine the discursive construction of race. Their sample study employs positioning theory and storyline analysis to examine racist online discourse. Jenny Hammond’s chapter on classroom discourse discusses the analysis of classroom talk. As a sample study, she chooses a project that examined a context in which English as a Second Language students were being inappropriately positioned as ‘failing students’ and ways in which they could be supported in the mainstream classroom. John Corbett’s discussion of discourse and intercultural communication commences with a definition of culture, then moves to distinctions between ‘cross-cultural’ and ‘intercultural’ communication. His study examines ways of managing rapport across cultures, in this case, relationships between English-speaking student language assistants and their mentors in French schools. The final chapter in the book on medical discourse by Timothy Halkowski examines the ways that ‘being a patient’, ‘being a doctor’, and indeed ‘being ill’, are accomplished in and through discourse. As can be seen, an expert in the particular aspect of discourse analysis has written each of the chapters of this book. Each author has provided a wealth of detail and given many examples of how his or her topic has been taken up in the research literature. We thank these authors for having agreed to contribute to this volume, and for sharing their expertise and experience in the ways that they have.

Further Reading Each of the chapters of this book provides suggestions for further reading on the particular topic being discussed and these are listed under Key Readings. 4

HylandK_Intro_Finals.indd 4

12/1/2010 8:10:59 AM

Introduction

Here we list some more general books on discourse analysis that readers will find useful as well. Bhatia, V. K., Flowerdew, J. and Jones, R. H. (eds) (2008), Advances in Discourse Studies. London: Routledge. Cameron, D. (2001), Working with Spoken Discourse. London: Sage. Gee, J. P. (2005), An Introduction to Discourse Analysis: Theory and Method (2nd edn). London: Routledge. Hoey, M. (2001), Textual Interaction: An Introduction to Written Text Analysis. London: Routledge. Jaworski, A. and Coupland, N. (eds) (2006), The Discourse Reader (2nd edn). London: Routledge. Johnstone, B. (2002), Discourse Analysis. Oxford: Blackwell. Paltridge, B. (2006), Discourse Analysis: An Introduction. London: Continuum.

5

HylandK_Intro_Finals.indd 5

12/1/2010 8:10:59 AM

HylandK_Intro_Finals.indd 6

12/1/2010 8:10:59 AM

Part I Methods of Analysis in Discourse Research

HylandK_Chap01_Finals.indd 7

12/1/2010 8:00:41 AM

HylandK_Chap01_Finals.indd 8

12/1/2010 8:00:41 AM

1

Data Collection and Transcription in Discourse Analysis Rodney H. Jones

Chapter Overview Data Collection as Mediated Action Five Processes of Entextualization Data in the Audio Age Video Killed the Discourse Analyst? Data Collection and Transcription in the Digital Age Conclusion Key Readings

9 10 11 16 18 19 20

Data Collection as Mediated Action The topic of this chapter is data collection and transcription, and, in it I will limit myself to the collection and transcription of data from real-time social interactions rather than considering issues around the collection of written texts, which has its own set of complications. Since the publication of Elinor Ochs’s groundbreaking 1979 article ‘Transcription as Theory’, it has become axiomatic to acknowledge that data collection and transcription are affected by the theoretical interests of the analyst, which inevitably determine which aspects of an interaction will be attended to and how they will be represented. In fact, this argument has been so thoroughly rehearsed by so many (see, for example, Bloom 1993, Edwards 1993, Mishler 1991) that there is little need to repeat it here. Neither do I intend to engage in debates about the ‘best system’ for transcribing spoken discourse (see, for example, Du Bois et al. 1993, Gumperz and

9

HylandK_Chap01_Finals.indd 9

12/1/2010 8:00:41 AM

Bloomsbury Companion to Discourse Analysis

Berenz 1993, Psathas and Anderson 1990) or ‘multimodal interaction’ (Baldry and Thibault 2006, Norris 2004), or about the need for standardization in transcription conventions (Bucholtz 2007, Lapadat and Lindsay 1999) since, to my mind, the acknowledgement that ‘transcription is theory’ basically pre-empts the need for such debates: if ‘transcription is theory’, one ought to be able to choose whatever system of representation best promotes one’s theory. I would like instead to focus on data collection and transcription as cultural practices of discourse analysts (Jaffe 2007), and examine how these cultural practices have changed over the years as different cultural tools (tape recorders, video cameras and computers) have become available to analysts, making new kinds of knowledge and new kinds of disciplinary identities possible. The theoretical framework through which I will be approaching these issues is mediated discourse analysis, a perspective which seeks to understand discourse through analysing the real time social actions it is used to take and the kinds of social identities and social relationships these actions make possible (Norris and Jones 2005). Central to this perspective is the concept of mediation, the idea that all actions, including the action of thought itself, are mediated through cultural tools (which include technological tools like tape recorders as well as semiotic tools like languages and transcription systems), and that the affordances and constraints of these tools help to determine what kinds of actions are possible in a given circumstance. This focus on mediation invites us to look at data collection and transcription not just as matters of theoretical debate, but as matters of physical actions that take place within a material world governed by a whole host of technological, semiotic and sociological affordances and constraints on what can be done and what can be thought, affordances and constraints that change as new cultural tools are introduced. Mediated discourse analysis, then, allows us to consider data collection and transcription as both situated practices, tied to particular times, places and material configurations of cultural tools, and community practices, tied to particular ‘kinds of people’ within particular disciplinary narratives.

Five Processes of Entextualization The primary cultural practice discourse analysts engage in is ‘entextualization’. We spend nearly all of our time transforming actions into texts and texts into actions. We turn ideas into research proposals, proposals into practices of interviewing, observation and recording, recordings into transcripts, transcripts into analyses, analyses into academic papers and academic papers into promotions. Ashmore and Reed (2000) argue that the business of an analyst consists of creating a series of artefacts – such as transcripts and articles – that are endowed with ‘analytic utility’. 10

HylandK_Chap01_Finals.indd 10

12/1/2010 8:00:42 AM

Data Collection and Transcription

Bauman and Briggs (1990) define ‘entextualization’ as the process whereby language becomes detachable from its original context of production and is thus reified as a ‘text’, a portable linguistic object. In the case of discourse analysts, this process usually involves two discrete instances of transformation, one in which discourse is ‘collected’ with the aid of some kind of recording device, and the other in which the recording is transformed into some kind of written or multimodal artefact suitable for analysis. Practices of entextualization have historically defined elite communities in society, who, through the ‘authority’ of their entextualizations are able to exercise power over others: scribes and story tellers, social workers and police officers, academics and lawmakers. To be engaged in creating texts about reality is to be engaged in creating reality. Whether we are talking about discourse analysts making transcripts or police officers issuing reports, entextualization normally involves at least six processes: (1) framing, in which borders are drawn around the phenomenon to be entextualized, (2) selecting, in which particular features of the phenomenon are selected to represent the phenomenon, (3) summarizing, in which we determine the level of detail with which to represent these features, (4) resemiotizing, in which we translate the phenomena from one set of semiotic materialities into another, and (5) positioning, in which we claim and impute social identities based on how we have performed the first four processes. These processes are themselves mediated through various ‘technologies of entextualization’ (Jones 2009), tools like tape recorders, video cameras, transcription systems and computer programs, each with its own set of affordances and constraints as to how much and what aspects of a phenomenon can be entextualized and what kinds of identities are implicated in this act. Changes in these ‘technologies of entextualization’ result in changes in the practice of entextualization itself, what it means, what can be done with it, what kinds of authority adheres to it, and what kinds of identities are made possible by it.

Data in the Audio Age The act of writing down what people say was probably pioneered as a research practice at the turn of the twentieth century by anthropologists and linguists working to document the phonological and grammatical patterns of ‘native’ languages. Up until 40 or so years ago, however, what people actually said was treated quite casually by the majority of social scientists, mostly because they lacked the technology to conveniently and accurately record it. On the spot transcriptions and field notes composed after the fact failed to offer the degree of detail necessary to analyse the moment by moment rhetorical unfolding of 11

HylandK_Chap01_Finals.indd 11

12/1/2010 8:00:42 AM

Bloomsbury Companion to Discourse Analysis

interaction. The ‘technologies of entextualization’ necessary to make what we now know as ‘discourse analysis’ possible were not yet available. This all changed in the 1960s when audiotaping technology became portable enough to enable the recording of interactions in the field. According to Erickson (2004), the fist known instance of recording ‘naturally occurring talk’ was reported by Soskin and John in 1963 and involved a tape recorder with a battery the size of an automobile battery placed into a rowboat occupied by two arguing newlyweds. By the end of the decade, the problem of battery size had been solved and small portable audio recorders became ubiquitous, as did studies of what came to be known as ‘naturally occurring talk’, a class of data which, ironically, did not exist before tape recorders were invented to capture it (Speer 2002). The development of portable audio-recording technology, along with the IBM Selectric typewriter, made the inception of fields like conversation analysis, interactional sociolinguistics and discursive psychology possible by making accessible to scrutiny the very features of interaction that would become the analytical objects of these fields. The transcription conventions analysts developed for these disciplines basically arose from what audiotapes allowed them to hear, and these affordances eventually became standardized as practices of ‘professional hearing’ (Ashmore et al. 2004) among certain communities of analysts. The introduction of these new technologies of entextualization brought a whole host of new affordances and constraints to how phenomena could be framed, what features could be selected for analysis, how these features could be represented and summarized, the ways meanings could be translated across modes, and the kinds of positions analysts could take up vis-à-vis others. Framing refers to the process through which a segment of interaction is selected for collection and/or transcription. Scollon and Scollon (2004) would doubtless prefer the term ‘circumferencing’ to ‘framing’. All data collection, they argue involves the analyst drawing a ‘circumference’ around phenomena, which, in effect, requires making a decision about the widest and narrowest ‘timescales’ upon which the action or interaction under consideration depends. All interactions are parts of longer timescale activities (e.g. relationships, life histories), and are made up of shorter scale activities (e.g. turns, thought units). The act of ‘circumferencing’, then, is one of determining which processes on which timescales are relevant to understanding what is ‘going on’. Among the most important ways audio recording transformed the process of framing for discourse analysts was that it enabled, and in some respects compelled them to focus on processes occurring on shorter timescales at the expense of those occurring on longer ones. One reason for this was that tapes themselves had a finite duration, and another reason was that audio recordings permitted the analyst to attend to smaller and smaller units of talk. 12

HylandK_Chap01_Finals.indd 12

12/1/2010 8:00:42 AM

Data Collection and Transcription

This narrowing of the circumference of analysis brought on by audiorecording technology had a similar effect on the processes of selecting and summarizing that went in to creating textual artefacts from recordings. Selecting and summarizing have to do with how we choose to represent the portion of a phenomenon around which we have drawn our boundaries. Selecting is the process of choosing what to include in our representation, and summarizing is the process of representing what we have selected in greater or lesser detail. The most obvious effect of audio-recording technology on the processes of selecting and summarizing was that, since audiotape only captured the auditory channel of the interaction, that was the only one available to the analyst for selection. Even though for many researchers the practice of tape recording was accompanied by the making of detailed observational notes regarding non-verbal behaviour, these notes could hardly compete with the richness, the accuracy, and the ‘authority’ of the recorded voice. As a result, speech came to be regarded as the ‘text’ – and all the other aspects of the interaction became the ‘context’. It is important to remember that this privileging of speech in our study of social interaction was not entirely the result of a considered theoretical debate, but also a matter of contingency. Analysts privileged that to which they had access. Sacks himself (1984: 26) admitted that the ‘single virtue’ of tape recordings is that they gave him something he could analyse. ‘The tape-recorded materials constituted a “good enough” record of what had happened,’ he wrote. ‘Other things, to be sure, happened, but at least what was on the tape had happened.’ Beyond limiting what could be selected to the audible, the technology of audio recording hardly simplified the selection process. Because tapes could be played over and over again and divided into smaller and smaller segments, the amount of detail about audible material that could be included in transcripts increased dramatically. Whereas most analysts based their decisions about what features of talk to include on specific theoretical projects – conversation analysts, for example, focusing on features which they believed contributed to the construction of ‘order’ in talk – some analysts, like Du Bois (Du Bois et al. 1993) promoted the development of more exhaustive systems of transcription, systems which not only fit present analytical interests but anticipated future ones. One thing for sure was that the dramatic increase in what could be included in textual representations of talk had the effect of making what discourse analysts did seem altogether more ‘scientific’, and over the years, the amount of detail in an analysts’ transcripts came to be seen as a criterion by which the ‘accuracy’ of their data and the ‘objectivity’ of their work was judged. As Mishler (1991: 206) describes it, researchers (strove) for more precision, detail, and comprehensiveness – pauses to be counted (by proper instruments) in hundreds rather than 13

HylandK_Chap01_Finals.indd 13

12/1/2010 8:00:42 AM

Bloomsbury Companion to Discourse Analysis

tenths of a second, the inclusion of intonation contours – as if that would permit us (finally) to truly represent speech. This desire to ‘truly’ reproduce speech was thoroughly grounded in positivist assumptions about reality – that there was something objectively occurring to reproduce – assumptions which would soon rub up against the more dialogic and constructionist theories that were arising from these very studies of talk in interaction (Scollon 2003). As transcripts revealed to analysts the contingent and negotiated nature of talk, analysts were themselves forced to confront the contingent and negotiated nature of their transcripts. More recently, analysts seem to be weighing in on the side of variety rather than standardization (Bucholtz 2007) and selectivity over comprehensiveness (Duranti 2006). Analysts like Jaffe (2007), in fact, have gone so far as to suggest that less delicate, more summative transcripts might in some cases constitute more ‘accurate’ representations of participants’ ‘voices’. Resemiotization is the process through which we translate phenomena from one set of semiotic materialities into another (Iedema 2001). Meanings are expressed differently in different semiotic systems, and so they cannot simply be transferred from one mode to another; they must be ‘translated.’ In data collection using audio recorders, for example, the social interaction, what is essentially a rich multimodal affair, is resemiotized into a mono-modal audiotape, later to be further resemiotized into a different mono-modal artefact, a written transcript. In this process, the spatial and temporal aspects of the dynamic, multimodal interaction must somehow be ‘translated’ into the static, linear and mono-modal materiality of text. One important aspect of resemiotization in written transcripts is how the spatial arrangement of elements on the page acts to translate certain temporal and relational aspects of the original interaction. Although there have been a number of experiments in the representation of interaction in written texts using non-standard layouts and notations (see, for example, Erickson 2003, Ochs 1979), most transcription systems developed for audiotaped data are arranged in the conventional ‘play-script’ layout, a layout that has a number of important effects on how we experience the interaction that is represented. First of all, the format creates the impression that interaction is focused, linear and mono-focal, masking any simultaneity of action, nonlinearity, or polyfocality that might have been part of the actual interaction. Second, it tends to imply a contingent relationship between immediately adjacent utterances of different speakers, whether or not one actually exists (Ochs 1979). Finally, it imposes on the transcript a particular ‘chronotrope’ (Bakhtin 1981) or felt ‘time-space’ that may be radically different from that of the original interaction. In fact, one of the most jarring discoveries of those coming fresh to discourse analysis is how much longer it takes to read through the transcript 14

HylandK_Chap01_Finals.indd 14

12/1/2010 8:00:43 AM

Data Collection and Transcription

of an exchange, with all of its details and pauses arranged linearly down the page, than it took the participants to actually produce the exchange in the first place. In short, the ‘play-script’ format requires that the reader rely primarily on the narrative interpretation of the analyst embodied in the sequential emplacement of lines and other elements on the page to make sense of what happened. Perhaps the most important process of entextualization, at least that with the most obvious social consequences, is positioning. Whenever we turn a phenomenon into a text, we are making claims as to who we are and what our relationship is to those whose words and actions we entextualize and those to whom we will later share these entextualizations. One rather obvious way that practices of data collection and transcription position the analyst is in how they reveal his or her affiliation to a particular ‘school’ of discourse analysis. It is, in fact, possible to give a cursory glance to the transcript in the appendix of an article and predict with a great deal of accuracy the kinds of theoretical positions about the language the writer will be advancing. As Jaffe (2007) has pointed out, transcription has become a kind of ‘literacy practice’, the mastery of which has become necessary for admittance into certain communities of scholars. Beyond signalling disciplinary affiliation, however, the new forms of transcription that audio recording made possible to discourse analysts also made possible for them new positions of authority in regard to their various audiences such as colleagues, tenure boards and funding bodies, as well as their ‘subjects’. This authority came, first, from the level of detail they were able to present in their transcripts, which they could use an emblem of ‘expertise’. Bucholtz (2000) has shown how the use of special fonts and annotations work to ‘technologize’ a text, and in the process, confer an identity of scientific expertise on the author. This new authority also came from the ‘evidentiary’ nature of the tape itself as a material object, the notion that by possessing the tape the discourse analyst had access to ‘what really happened’ against which both the ‘authenticity’ of the transcript and any claims or counter-claims about it could be measured. Ashmore and his colleagues (2004) call the tendency to confer on ‘the tape’ an epistemic authority ‘tape fetishism’. The most dangerous thing about such an attitude is not just that the supposed ‘authority’ and ‘objectivity’ of ‘the tape’, produced as it was in particular circumstances of recording and listened to in varying contexts of hearing, are so easily called into question, but also because the existence of the tape itself lends further authority to the transcript, which is presumed to be the ‘child’ of the tape. This overconfidence in tapes and transcripts in the domain of discourse analysis simply makes for sloppy work. In other domains like law enforcement (Bucholtz 2009) the consequences can be rather more serious. 15

HylandK_Chap01_Finals.indd 15

12/1/2010 8:00:43 AM

Bloomsbury Companion to Discourse Analysis

With the new authority granted to discourse analysis by the invention of the portable tape recorder, there also came new responsibilities. For one thing, analysts found themselves embedded in a complex new set of ethical and legal relationships with the subjects of their analysis. Much of the pioneering work using audio recorders simply ignored this complexity – it is hard to imagine, for example, how Sacks’s recording of suicide hotlines would be treated in light of today’s standards of ‘informed consent’. Eventually, however, ethics boards and the law caught up with us. Not only do institutional review boards now demand that informed consent be obtained from any party whose voice is tape-recorded, but in many countries the law also demands it. These constraints, have left the discourse analyst struggling to find ways to preserve the ‘naturalness’ of interactions in which all involved are aware they are being taped, a most ‘unnatural’ state of affairs. The great irony of recording technology for discourse analysts is that it simultaneously introduced a standard of ‘naturalness’ for our data and created social and institutional conditions that made that standard much more difficult to obtain.

Video Killed the Discourse Analyst? Audio recording was not the only technology social scientists used in the midtwentieth century to study communication. As early as the 1940s, Gregory Bateson and Margaret Mead (1942) were pioneering the use of film in the study of human communication, a technique that was later taken up by Edward Hall (1963) in his early studies of proxemics. By the 1970s, analysts like Birdwhistell (1970) had begun to develop transcription systems for non-verbal features of social interaction. The assumption of these analysts was that meaningful interaction proceeds not just through talk, but through a whole host of other behaviours as well. This assumption would nowadays be considered non-controversial, but, in the heady days of tape recording of the 1960s and 70s, it failed to gain much traction, not until, of course, the invention of the video camera, a new ‘technology of entextualization’ capable of capturing not just words but also bodies in motion in a much cheaper and more immediate way than earlier film technology. Discourse analysis was ruined forever. Only in one sense, that is. Discourse analysts could no longer ignore nonverbal behaviour, which played so demonstrably an important role in most social interactions. And the technology that allowed analysts access to that behaviour involved a whole new set of processes through which discourse analysts could frame, select, summarize and resemiotize their data and position themselves in relation to it. One important change came with the analyst now being able to frame his or her data spatially as well as temporally. With audiotape, only the duration, the 16

HylandK_Chap01_Finals.indd 16

12/1/2010 8:00:43 AM

Data Collection and Transcription

starting point and ending point of the interaction mattered. Now the interaction had to be framed in space as well, with a whole new set of choices to be made about what and who should be included in the frame, the angle at which it should be shot, and so forth. Video also made the choices involved in selecting and summarizing much more complex, as nearly every aspect of non-verbal communication from gesture to gaze to body movement and posture could be considered potentially communicative, as could a whole host of other non-verbal cues like dress and built environment. The biggest difficulty, however, came in the process of resemiotization, the challenge of translating the rich, multidimensional display of videotape to the still-dominant two dimensional medium of the written transcript (Park and Bucholtz 2009). Early users of video essentially treated it as an extension of the audio recorder, using it as an aid to adding information about such things as gesture and gaze as notations within what were essentially conventional audio transcriptions (see, for example, Goodwin 1986, Ochs and Taylor 1992). Many early attempts at multimodal transcription, were hindered by the essentially ‘verbal logic’ of the ‘play script’ model which analysts had inherited from the audio days, a model which provided few resources for representing the complex timing and simultaneity of actions and words in multimodal interaction. The problem with most early work using video was that technologies of transcription had not yet caught up with the technologies of recording. At the same time, video introduced further complexity into the analyst’s relationships with other people. Given the fact that video data so clearly identify their objects, it became much more difficult to promise anonymity and confidentiality to participants, and consequently, more difficult to obtain ‘informed consent’ (Yakura 2004). Furthermore, the ‘gaze’ of the camera in many ways turned out to be much more intrusive than the ‘ear’ of the tape recorder, giving rise to new layers of self-consciousness and artificiality compromising the ‘naturally occurring’ status of our data. Video technology also had an effect on the analyst’s relationship with the consumers of his or her data, particularly the publishers of academic journals and books who were in those early years still reluctant to go to the extra expense of publishing the photographs and other visual data many analysts found essential for communicating their findings, and the print medium itself, still the only medium that seemed to garner any recognition from academic institutions, lacked the ability to give readers access to anything but static images. In the 1980s, the constraints and complications of video recording often seemed to outweigh the dramatic new affordances the medium offered to analysts, and many, despite overwhelming evidence of the importance of the visual as well as the verbal in social interaction, held stubbornly to the monomodal talk-based approach to interaction which had served them so well in the past. This, however, was soon to change. 17

HylandK_Chap01_Finals.indd 17

12/1/2010 8:00:43 AM

Bloomsbury Companion to Discourse Analysis

Data Collection and Transcription in the Digital Age Many of the issues that plagued early users of video began to be resolved at the turn of the century as analysts like Baldry and Thibault (2006) and Norris (2004) began devising fully theorized systems of multimodal transcription. These breakthroughs, however, were not all theoretical. They came as well from another dramatic material change in the ‘technologies of entextualization’ available to the analyst, a change that was made possible by the digital revolution. The qualitative difference between analog recording and digital recording as technologies of entextualization cannot be overstated. Digital audio and video can be handled and manipulated by the analyst in so many different ways, many of which are reminiscent of the ways we handle and manipulate written text – it can be searched, tagged, annotated, chopped up, rearranged and mixed with other texts. Because of this, it creates a whole host of new affordances when it comes to transcription. Digital video did not so much change how analysts were able to record video as it did what they were able to do with it afterwards, the new means they had at their disposal to create objects of ‘analytic utility’ out of it. The ability to easily capture still images from video meant that analysts no longer had to rely solely on text to describe behaviour. Text and images could be integrated in ways that made transcripts themselves ‘multimodal’. The practice of including still images captured from digital video in transcripts has been developed to great sophistication by scholars like Baldry and Tibault (2006) and Norris (2004). Such ‘multimodal transcripts’, however, are still not the most multimodal means we have at our disposal to represent our data. Gu (2006), for example, has promoted the use of a ‘corpus friendly’ digital multimedia system for representing interaction which avoids the need for orthographic transcription altogether, and software solutions like Transana and Elan allow analysts to integrate their videos with their transcripts, their coding and their notations in flexible, searchable ways (Mondada 2009). Such advances have given to analysts the feeling that they are closer to the ‘reality’ of the original interaction than ever before. But they are not. They are closer to a digital fabrication of reality, which is still only a fabrication. Just as a kind of ‘tape fetishism’ led analysts in the audio age to believe that ‘accurate’ and exhaustive transcripts of tape recordings would allow them to once and for all truly represent speech, the ability these new digital solutions give to analysts today to analyse video ‘directly’, seemingly unmediated by the transcription process, creates the illusion that they do not have to truly represent anything (that the video has done that for

18

HylandK_Chap01_Finals.indd 18

12/1/2010 8:00:43 AM

Data Collection and Transcription

them), the illusion that the complex problems of selecting and summarizing can be somehow sidestepped, and that the inevitable distortions that accompanied the transformation of audiotape to written text can now be completely avoided, in short, the illusion that the age of ‘transcriptionless analysis’ has arrived. As Mondada (2009) has pointed out, however, the viewing, coding and otherwise manipulating of video data with such software packages is far from unmediated. Users still need to go through the same five processes of entextualization that transcribers apply to audiotape. They still need to determine what counts for them as a meaningful unit of social interaction; aspects of the data still need to be selected, coded or otherwise summarized; and videos are still resemiotized into complex ‘semiotic aggregates’ combining symbols and writing with the audio and visual modes of the video. Unlike written transcripts, multimodal texts have no ready-made ‘textual units’ apart from time codes, and so analysts must engage themselves in inventing new ways to divide up the dynamic stream of behaviour into manageable, intelligible bits. And these products of entextualization need to be still further entextualized into objects that can be published in an academic press still dominated by the medium of print. Software can impose just as sturdy a set of theoretical assumptions on the analyst as a transcription system. All entextualizations are necessarily arrived at dialogically and are thus inherently ‘double-voiced’ (Bakhtin 1984: 185). By loosing sight of this ‘double-voicedness’, by thinking they can ‘sidestep’ the gap between the original and the entextualized, discourse analysts are in danger of regarding their ‘multimodal transcripts’ as somehow more objective and transparent, of believing that we have finally found a way to truly represent social interaction. The notion that a ‘multimodal transcription’ of a video is necessarily a more ‘accurate’ portrayal of ‘reality’ than a careful transcription of an audiotape is really a matter of opinion. It depends primarily on how one defines ‘reality’. Both annotated video and written transcripts are artefacts, products of complex processes of framing, selection, summarizing and resemiotization, whose meanings change as they are transported across boundaries of time, space, contexts and media (Jaffe 2007).

Conclusion In his article ‘The Dialogist in a Positivist World’, Scollon (2003) explores the balancing act discourse analysts have to perform to avoid, on the one hand, over-reifying their data and falling into a naive positivism, and, on the other

19

HylandK_Chap01_Finals.indd 19

12/1/2010 8:00:43 AM

Bloomsbury Companion to Discourse Analysis

hand over-relativizing their data and sinking into deconstructive impotence. The chief concern for a discourse analyst, he argues, is how to ‘produce a working ontology and epistemology that will underpin (his or her) wish to undertake social action’ without buying into the social constructions that underpin this action (p. 71). The sometimes-paradoxical history of data collection and transcription in discourse analysis presented here is really the history of this dilemma. And I suppose the lesson of this history for anyone starting out in discourse analysis is that no technology of entextualization can capture the universe (Cook 1990). Nor is this what we need. The whole reason for entextualization is not to reproduce the universe, but to re-present it, and, by doing so, to understand it better. And it is through these very processes of framing, selecting, summarizing, resemiotizing and positioning that which we arrive at these understandings. Too often analysts have taken a ‘deficit’ attitude towards entextualization, lamenting how much of the ‘original’ interaction was ‘lost in the transcription’. The fact is what we search for in our transcripts is not ‘truth’, but rather ‘analytic utility’. Their ability to help us answer the questions we have about human communication and social interaction, not the degree to which they ‘resemble reality’, should be the main criterion for judging the value of our transcripts. At the same time, we must never lose sight of the ways technologies of entextualization profoundly affect our relationships with those whose words and behaviour we study. The better our technology has become at capturing the details of social interaction, the more pressing and complex have become the ethical issues surrounding the activities of data collection and transcription. As Scollon and Levine (2004: 5) write: The primary question now is not: Do we have or can we develop the technology needed to record the behavior of others? The primary question is: What rights does an academic researcher have in relationship to and in negotiation with her or his subjects of study? . . . In short, can our data collection and our analyses do others good or harm, and can we control those outcomes?

Key Readings Duranti, A. (2006), ‘Transcripts, like shadows on a wall’, Mind, Culture and Activity, 13, 301–10. Edwards, J. A. and Lampert M. D. (eds) (1993), Talking Data: Transcription and Coding in Discourse Research. Hillsdale, NJ: Lawrence Erlbaum.

20

HylandK_Chap01_Finals.indd 20

12/1/2010 8:00:44 AM

Data Collection and Transcription Lapadat, J. C. and Lindsay, A. C. (1999), ‘Transcription in research and practice: from standardization of technique to interpretive positionings’, Qualitative Inquiry, 5, 64–86. Norris, S. (2002). ‘Implications of visual research for discourse analysis: transcription beyond language’, Visual Communication, 1, 97–121. Ochs, E. (1979), ‘Transcription as theory’, in E. Ochs and B. Schieffelin (eds), Developmental Pragmatics. New York: Academic Press, pp. 43–72.

21

HylandK_Chap01_Finals.indd 21

12/1/2010 8:00:44 AM

2

Conversation Analysis Sue Wilkinson and Celia Kitzinger

Chapter Overview Introduction: Origins and Key Features Data Analysis: The ‘Building Blocks’ of Social Life The Turn-taking System Collaborative Completion Summary, Conclusion and Future Prospects Transcription Key Note Key Readings

22 24 28 31 34 36 37 37

Introduction: Origins and Key Features Conversation analysis (CA) – the study of talk-in-interaction – is a theoretically and methodologically distinctive approach to understanding social life. It is now an interdisciplinary field – spanning, in particular, sociology, psychology, linguistics and communication studies. It was first developed within sociology in the United States in the late 1960s and early 1970s by Harvey Sacks and his collaborators, Emanuel Schegloff and Gail Jefferson (e.g. Sacks et al. 1974). Sacks studied law before turning to sociology, subsequently moving to California to work with Harold Garfinkel. His thinking was influenced by the then-developing sociological tradition of ethnomethodology, which looks at people’s ways of making sense of the everyday social world, and drew especially on Garfinkel’s early work as well as that of Erving Goffman. These origins underpin CA’s understanding of talk as, first and foremost, a form of action, and its focus on discovering what people do with talk in the course of their everyday lives (rather than just on what they say). Tragically, Sacks was killed in a car crash in 1975, leaving much of the subsequent development of CA to his collaborators, colleagues and students. His 22

HylandK_Chap02_Finals.indd 22

12/1/2010 8:01:08 AM

Conversation Analysis

foundational legacy remains – in its written form principally as lectures to his undergraduate students at the University of California, transcribed from the original tapes by Jefferson, and published with an extensive introduction by Schegloff (Sacks 1995). Sacks did not set out to study conversation, but it so happened that, as a Fellow at the Center for the Scientific Study of Suicide in Los Angeles in 1963, he was given access to recordings of telephone calls to the Suicide Prevention Center. He analysed what happened in these calls – including accounts of suicide threats getting laughed off, and descriptions of suicide as attempts to ‘discover if anyone cares’. But Sacks’s focus was not primarily on the topic of suicide. Rather, he found that even calls to a Suicide Prevention Center are full of the kinds of activities most of us do every day: requesting, agreeing, disagreeing, and so on – and he set about analysing in detail how these everyday activities get done. Over the 50 years or so since Sacks’s pioneering work, a large body of empirical work has uncovered the key structural features of talk-in-interaction – the ‘building blocks’ upon which all social life depends. These include: how people get to take turns at talk, how actions are organized into sequences, and how speakers and listeners deal with trouble in speaking, hearing or understanding the talk. We will say more about these ‘building blocks’ in the next section. Conversation analysts use these empirical findings in studying both ordinary conversational actions, such as complaining (Drew and Holt 1988), complimenting (Pomerantz 1978), or telling news (Maynard 1997); and also actions that constitute particular ‘institutional’ contexts, such as classrooms, courts or call centres (Drew and Heritage 1992), medical consultations (Heritage and Maynard 2006), or telephone helplines (Baker et al. 2005). Two features of Sacks’s original data set (the suicide helpline calls) remain central to CA work today. First, data are naturally occurring, rather than researcher-generated. Sacks did not interview suicidal people about their experience of calling the helpline – he analysed the helpline calls themselves. He analysed an actual piece of social life, rather than a retrospective report of it. This meant that the helpline callers were pursuing actions in their own lives – addressing their own concerns, rather than answering research questions. CA today is still based on the analysis of actual instances of talk-in-interaction (rather than invented or hypothetical instances or retrospective self-reports generated by researchers via interviews or focus groups). Second, data are recorded, allowing repeated playback for analysis. The invention of the tape recorder made this possible for Sacks: the availability of recordings enabled a much more detailed inspection of talk-in-interaction than had previously been possible (e.g. attending to features of timing, such as silences or overlapping talk; and features of delivery, such as hesitation or emphasis). Sacks found – and subsequent work has confirmed – that interactional participants 23

HylandK_Chap02_Finals.indd 23

12/1/2010 8:01:09 AM

Bloomsbury Companion to Discourse Analysis

are oriented to such fine-grained details of talk-in-interaction, and that these are highly consequential for how an interaction develops. Contemporary CA uses audio- or video-recorded data. Video is necessary for the study of face-toface interactions, and enables an analysis of how interactional features such as gesture, body deployment and gaze are used (e.g. Goodwin 1981). Transcription of data for CA analysis uses detailed notation invented by Gail Jefferson (Jefferson 2004) representing various features of the timing and delivery of talk. In Extract 1 below, for example, the underlining on the first part of the word early (line 1) indicates emphasis, while the colon indicates that the preceding sound is stretched; the period in parentheses (line 2) represents a tiny (but nonetheless hearable) silence of less than a tenth of a second; and the dash following the w (line 5) indicates that the sound is cut off in the course of its production. There is a full transcription key at the end of this chapter. It is the recordings (not the transcripts) that constitute the primary data of CA, and these are regularly revisited (and transcripts regularly revised) during the analytic process. CA is unusual in that data sets are often shared – indeed, many of the ‘classic’ data sets are available as a resource for the CA community. They are widely used in teaching, frequently re-analysed for new phenomena, and appear in publications by a range of different authors. Best practice in contemporary CA – subject to obtaining the appropriate permissions – is to make the data extracts displayed in published articles available on the web. For example, you can listen to many extracts from Emanuel Schegloff’s articles at www.sscnet.ucla.edu/soc/faculty/schegloff; and some of the data in this chapter can be heard at www.lboro.ac.uk/departments/ss/staff/ wilkinson.html.

Data Analysis: The ‘Building Blocks’ of Social Life Each episode of talk-in-interaction has unique characteristics. It takes place under particular circumstances between particular individuals, living in specific social and cultural contexts, who bring to the interaction their own personal characteristics, experiences and beliefs, as well as their relationship history. But each is also made up of some recurrent features – what we have referred to as the ‘building blocks’ of social life. CA is concerned with identifying these features and understanding how they are used in action. Data analysis, then, focuses on the organized, recurrent structural features of talk-in-interaction, features which stand independently of the characteristics of individual speakers or the relationship between them. Knowledge of these structural features (or, in Schegloff’s [2007] terminology, ‘generic orders of organization’) is a key part of our competence as social actors: it influences both our own conduct in talk-in-interaction and our interpretation of the conduct of others. 24

HylandK_Chap02_Finals.indd 24

12/1/2010 8:01:09 AM

Conversation Analysis

CA offers technical specifications of six key structural features of talk-ininteraction: turn-taking, action formation, sequence organization, repair, word selection and the overall structural organization of talk. We will sketch out each of these in relation to a single brief data extract, and then focus in more detail on just one of them: the turn-taking system. Extract 1 comes from a phone conversation between a married couple (Edna and Bud), on their wedding anniversary. Edna is already at the couple’s vacation home at the beach and it is apparent from what she says earlier in the call that she had expected Bud to join her there the previous day, in time for their anniversary. The call begins with exchanges of ‘Happy Anniversary’, later reiterated by Edna, who tells her husband ‘I miss you’. As this extract opens, Edna is asking Bud about his travel plans for the following day, and is clearly hoping he will be joining her soon. As it turns out, her hopes are to be frustrated: Extract 1 [NB III:3]1 01 Edn: You comin’ down ea:rly? 02 (.) 03 Bud: We:ll I got a lot of things to do before 04 I get cleared up tomorrow. >I dunno.< 05 I w- I probably won’t be too early. Below, we examine this short extract in relation to the six ‘building blocks’ of talk-in-interaction.

Turn-taking We can notice that this data extract consists of two turns at talk from different speakers. They speak one at a time, and there is a brief silence between speakers. Their turns are different lengths: the first is a single unit of talk; the second is three distinct units. The classic paper by Sacks, Schegloff and Jefferson (1974) presents a model of turn-taking which accounts for such observations (and more). We will describe the turn-taking system in more detail in the next section.

Action Formation Action formation refers to the ways in which speakers fashion turns to be recognizable to their recipients as doing a particular action (e.g. complaining, inviting, declining, and so on). One way of characterizing the two turns in 25

HylandK_Chap02_Finals.indd 25

12/1/2010 8:01:09 AM

Bloomsbury Companion to Discourse Analysis

Extract 1 is to say that Edna asks a question and Bud answers it. We can notice, for example, how Edna uses prosody – rather than grammar – as a resource to frame up her action as a question: the rising intonation on early, indicated by the question mark, is a way of ‘doing questioning’ (where a falling intonation might have led it to be heard as a command). Bud shows that he is answering his wife’s question in part by repeating elements of it: for example, the word early.

Sequence Organization It is common to find, as in Extract 1, that a question is followed by an answer – or that an invitation is followed by an acceptance/declination, a news announcement by a news receipt, and so on. Actions are organized into sequences, the most basic type of which is the ‘adjacency pair’: two turns at talk by different speakers, the first constituting an initiating action, and the second an action responsive to it. Most initiating actions (also called ‘first pair parts’ or FPPs) can be followed by a range of appropriately ‘fitted’ next actions (‘second pair parts’, or SPPs). Some SPPs further the action of the prior FPP (e.g. accepting an invitation) and are termed preferred responses, others do not (e.g. rejecting an invitation) and are termed dispreferred. In Extract 1, Edna’s FPP ‘prefers’ the answer ‘yes’ (i.e. confirmation that Bud will be coming down early) – and you will notice that this is not the answer it gets. In effect, Bud is saying that he will not be coming down early (a dispreferred response) – and his SPP has many of the features that CA has found are characteristic of dispreferred SPPs. Whereas preferred responses tend to occur without delay, and to be short and to the point, dispreferred responses are likely to be delayed and elaborated – as here, where Bud’s dispreferred response is delayed first by a short silence, and then by a turn-initial marker (well), a hedge (I dunno) and an account (I got a lot of things to do . . . ) before he actually answers the question. A version of this data extract and analysis appears in Schegloff (2007: 68–9): his book is the authoritative work on sequence organization.

Repair It is quite common for speakers to treat what they are saying as problematic in some way and to stop what they are saying in order to fix the problem. So in Extract 1, Bud cuts off his talk at a point where it cannot be possibly complete (after I w-, line 5), in order to go back and add something (the word probably) into his turn-in-progress – technically, an insertion repair (Wilkinson and Weatherall in press). Insertion repair (like other repair practices) can be used 26

HylandK_Chap02_Finals.indd 26

12/1/2010 8:01:10 AM

Conversation Analysis

to accomplish a variety of actions: here Bud seems to be softening the blow of responding to his wife’s hopeful enquiry with a dispreferred SPP (i.e. telling Edna that he won’t be at their beach home as early as she would like). Other ways in which speakers repair their own talk include, for example, replacement (replacing one word or phrase with another), deletion (removing a word from a turn-in-progress), and reformatting (Schegloff et al. 1977). Recipients of a turn can also initiate repair on it if they find it problematic in some way (e.g. pardon? or huh? may be used to claim a problem of hearing): the classic reference on other-initiated repair is Schegloff (2000a).

Word Selection Turns at talk are composed of lexical items selected from amongst alternatives. For example, we have seen that Bud selects the word early – first used by Edna – as one way of showing that he is answering her question. We can also notice that he selects the formulation a lot of things to do – suggesting time-consuming activity – in accounting for why he won’t be coming down early. CA explores how word selection is done as part of turn design and how it informs and shapes the understanding achieved by the turn’s recipient. It has focused particularly on category-based ways of referring to non-present persons: for example, law enforcement officers can be referred to as ‘police’ or as ‘cops’, and speakers’ selection of one or the other may be responsive to whether the speaker is appearing in court (Jefferson 1974) or talking with adolescent peers (Sacks 1995).

Overall Structural Organization Talk-in-interaction is organized into phases: most obviously, openings and closings (Schegloff and Sacks 1973). Extract 1 comes from a call launched (by Edna) with a ‘Happy Anniversary’ sequence – and the beginning of a call is the proper place to exchange best wishes for anniversaries, birthdays etc., or to register other ‘noticeables’ (Schegloff 2007: 86–7). Invocations of future interaction – such as see you Thursday or I’ll be in touch then – are common at the end of calls. In institutional interactions there are often component phases or activities which characteristically emerge in a particular order. Acute doctor– patient interactions, for example, have a highly structured overall organization: Opening, Presenting Complaint, Examination, Diagnosis, Treatment and Closing (Heritage and Maynard 2006) – and doctors’ and patients’ conduct can be analysed for how they orient to and negotiate the boundaries of each phase of the interaction. Many recent studies draw on analyses of overall structural 27

HylandK_Chap02_Finals.indd 27

12/1/2010 8:01:10 AM

Bloomsbury Companion to Discourse Analysis

organization as part of research designed to be of practical use by organizations in improving the quality of their services (see Antaki forthcoming). We will now look in more detail at just one of these structural features of talk-in-interaction: the turn-taking system.

The Turn-taking System The organization of turn-taking is fundamental to talk in interaction: understanding how turn-taking works is an essential prerequisite for CA research. A classic foundational CA article (Sacks et al. 1974) lays out a model for turntaking in conversation, which has subsequently been developed by these and other researchers (e.g. Jefferson 1986; Lerner 1991, 1996; Schegloff 2000b, 2001). Although turn-taking is sometimes thought of as a purely technical phenomenon, it has, as we will show, hugely significant interactional implications. Understanding how turn-taking normally works in conversation is important for analysts both because co-conversationalists use the turn-taking system to pass the conversational floor between them in an orderly way, and because participants can manipulate this normative system to bring off particular interactional effects – including displays of power, (non-)cooperation or empathy. The key observation of Sacks, Schegloff and Jefferson’s (1974) classic paper was that normally (with orderly and describable exceptions: for example, Lerner 2002) people speak one at a time in conversation. Their paper sets out to explain how this is achieved using the concept of ‘projectability’. By ‘projectability’ they mean that any unit of talk can be monitored, in the course of its production, for what it will take to bring it to possible completion (e.g. to finish the sentence). Co-interactants track their own talk and that of their coconversationalist(s) in the course of its production, using syntax, prosody and pragmatics as resources to project when a turn is coming to possible completion. A next speaker may legitimately start to speak at a place where a turn at talk is possibly complete: a so-called transition relevance place (TRP). This is typically at the end of the first ‘turn constructional unit’ (TCU): that is the first sentence, clause, phrase, or lexical item out of which a turn can (in any given context) be constructed. A recipient of a turn at talk, awaiting their turn to talk next, listens to the talk-in-progress in part to project (or predict) when the unit of talk will be done, in order that they can talk next with no gap and no overlap. Transitions from one turn to a next without gaps or overlaps make up the vast majority of transitions in conversation and when this happens the turn-taking system is working normally. Variations on this normal transition are often interactionally consequential. We have already seen (in Extract 1) that a gap of silence between one turn

28

HylandK_Chap02_Finals.indd 28

12/1/2010 8:01:10 AM

Conversation Analysis

and the next can be used to foreshadow a ‘dispreferred’ response (i.e. disappointing a hopeful expectation). Speakers of initiating actions hear delays in responsive turns as dispreference-implicative, and show their understanding by, for example, attempting to deal with possible obstacles to acceptance. In Extract 2, Fran is speaking with Ted about arrangements for their daughters to get together. When Fran says she will have to drive her daughter over to Ted’s beach house, Ted issues an invitation for Fran to stay too: Extract 2 [NB.III.1] 01 Ted: mWell y’c’n both sta:y. hh 02 (0.2) 03 Fra: [Oh uhh ] 04 Ted: [Got ple]nty a’roo:m, hh[hh Fran’s silence at line 2 occurs at a place where turn transition is relevant since Ted’s turn is hearably complete pragmatically (it does a complete action – an invitation), grammatically (it is a complete sentence), and prosodically (the elongated vowel sound on the last word and the falling intonation are both typical of prosodically complete TCUs). Analysing the silence on line 2, we can use the turn-taking model to say who it is who is not speaking (Fran), and sequence organization to say what it is that she is not saying (accepting or declining, since these are the relevant SPPs following a FPP invitation). Since CA analyses of preference organization suggest that silence at the transition relevance place foreshadows a dispreferred SPP, we can presume that a rejection is in the works. Crucially, this is not just our view as academic analysts – Ted hears it that way too. He shows he hears the silence as dispreference-implicative by speaking again (just as Fran also finally starts to respond) to address a possible obstacle: that Fran might think the beach house too small to accommodate her as well as her daughter. (As it turns out, Fran declines the invitation on the different grounds that she has responsibilities to her two other children.) We have seen, then, that delayed turn transition – that is leaving a silence after a TCU has been brought to possible completion in the slot where its recipient should speak next – can be interactionally consequential. Taking a turn prematurely is also fraught with interactional meanings. ‘Interrupting’ means launching a turn at a point where a speaker’s TCU is not possibly complete. An important finding of CA research – indeed, one that underpins the turn-taking model itself – is that in most overlapping talk speakers start up at a place where a turn could have been possibly complete, but – as it turns out – is not, as in Extracts 3–4 (the square brackets, one above the other on consecutive lines, indicate overlap):

29

HylandK_Chap02_Finals.indd 29

12/1/2010 8:01:10 AM

Bloomsbury Companion to Discourse Analysis

Extract 3 [from Sacks, Schegloff and Jefferson, 1974: 708] 01 A: What’s yer name again please [sir, 02 B: [F.T. Galloway. Extract 4 [from Sacks, Schegloff and Jefferson, 1974: 707] 01 A: Uh you been down here before [havenche. 02 B: [Yeh. In both instances, Speaker A has produced as part of their turn what Sacks, Schegloff and Jefferson (1974) term ‘optional elements’: the address term ‘sir’ (Extract 3), the tag question ‘haven’t you’ (Extract 4). In both instances, Speaker B started up at a point at which it was projectable that the prior turn would be complete and that their turn would be launched in the clear without either a gap or an overlap. Because of the ‘optional’ element included in the prior turn, this turned out to be a misprojection. The second speaker, in effect, ‘mistimed’ their entry into the talk. Brief and unproblematic instances of overlap like these are very common in conversation and they are characteristically resolved by one speaker stopping talking a few syllables (or beats) into the overlap (Schegloff 2000b, 2001). Relatively little overlap results because a recipient manifestly ‘interrupts’ a current speaker, that is launching their own turn at a place at which the ongoing turn cannot be possibly complete. Extract 5 shows one such example taken from a telephone conversation between Edna (the same Edna as in Extract 1) whose husband, Bud, has just left her. She is complaining about him to her sister Lottie. Extract 5 [NB IV:10] 01 Edn: He c’n make me so da:mn mad I c’d 02 (.) [bop him] °but° ] 03 Lot: [uW’l tha]t’s a’]way with me[↓:.]too↓:.] The silence at the beginning of line 2 is not at a place where Edna could have finished her turn. The TCU that begins He can make me so damn mad I could is not possibly complete grammatically: projectable possible completions are (for example) verbs such as cry or scream, or verb phrases such as burst into tears, or (what she eventually says) bop him (i.e. hit him). In other words, although Edna is not speaking at line 2, she still ‘owns’ the turn: this is a mid-TCU silence (a ‘pause’) and not a silence at a transition relevance place (a ‘gap’). Nonetheless, 30

HylandK_Chap02_Finals.indd 30

12/1/2010 8:01:10 AM

Conversation Analysis

Lottie starts to speak at this place – and finds herself in overlap with Edna’s continuation. Notice, however, that Lottie’s ‘interruption’ is not hostile or aggressive. Rather the contrary, she is aligning with Edna’s position on the provoking nature of husbands. It turns out that most instances of ‘interruption’ – that is where a next speaker begins in the course of a current speaker’s turn and not close to a transition relevance place – are affiliative rather than hostile (Jefferson 1986; Schegloff 2000b, 2001). In this section, then, we have seen how normative turn-taking rules – that speakers should start to talk at the end of a previous speaker’s TCU without gap or overlap – are breached in ways that are interactionally consequential. As analysts we need to be able to analyse turns at talk for where a TCU is possibly complete, what it would take to bring it to possible completion, who is not speaking in a silence, whether an incoming turn is ‘mistimed’ or ‘interruptive’, whether an ‘interruption’ is affiliative or hostile, and so on. We need to be able to analyse this because participants are producing these analyses for themselves continuously in the course of their interactions.

Collaborative Completion We have seen some of the ways in which the turn-taking system constitutes a resource for interactional participants, and – particularly – how breaches of the turn-taking ‘rules’ have particular interactional import. We turn now to a more extended example of such a breach – the practice of ‘collaborative completion’ – and its interactional significance. The turn-taking model shows that normally people do not take a turn at talk until the TCU-in-progress produced by their co-conversationalist reaches possible completion. But in Extracts 6 and 7 below, speakers are clearly part way through a TCU when their co-conversationalist starts to talk (look at lines 4 and 5 in Extract 6; and lines 6 and 7 in Extract 7). You might notice, however, that the person who comes in mid-TCU is doing something very special: they are not producing a new turn of their own (as Lottie does in Extract 5), but finishing the turn their co-conversationalist was in the course of producing. Extract 6, taken from a conversation between a group of friends drinking beer around a picnic table in the garden of one couple’s home, involves a complaint about a neighbour (the she of line 2). A single TCU is jointly constructed by Chris (who begins it) and Carol (who ends it). Extract 6 [AutoDiscussion] 01 Chris: °God damn it,° hh[hhh! ] 02 Carol: [What’d] she do:¿ 31

HylandK_Chap02_Finals.indd 31

12/1/2010 8:01:11 AM

Bloomsbury Companion to Discourse Analysis

03 (0.4) 04 Chris: Oh: nothin’ she’s just a- (0.2) 05 Carol: bi:tch. 06 Chris: tch! Ye:ah. Although Chris starts the complaint (she’s just a-, line 4), he doesn’t complete it. Instead, after producing – and cutting off – the word a he leaves two-tenths of a second of silence. It is Carol who completes the TCU that Chris has begun, adding the word bitch (line 5), which Chris accepts as correct (Yeah, line 6). Extract 7 is taken from a telephone conversation between two nurses. They are discussing a third nurse (the she of lines 1 and 5) whom Bee is suggesting as a possible relief nurse – someone whom Amy feels unable to recommend for the job. Again, a single TCU is jointly constructed: it is started by Amy and ended by Bee. Extract 7 [SBL] 01 Bee: Well do you think she would fit i:n¿ 02 (.) 03 Amy: Uh:m .hh (.) uh I don’t kno:w=What I’m: 04 hesitating abou:t is uhm .hh uhm (.) maybe 05 she wou:ld .hh (0.8) uh but I: would 06 hesitate to: uhm 07 Bee: recomm[en]d her. 08 Amy: [I-] 09 Amy: Ye:s. Although Amy starts the negative assessment (I would hesitate to, lines 5–6), she doesn’t complete it. Instead, after the word to she produces a filled pause (uhm) and it is Bee who completes the TCU that Amy begun, adding the words recommend her (line 7), which Amy accepts as correct (Yes, line 9). As in Extract 6, a single TCU is collaboratively produced by two people. Collaborative completion (or, as it is sometimes termed, ‘anticipatory completion’) is a particularly striking phenomenon, in which a single TCU is jointly constructed by two different speakers: the first speaker begins the TCU and the second speaker ends it. This phenomenon has been extensively studied by Gene Lerner (Lerner 1991, 1996a, b, 2004, in press), who explores how it is possible for people to do collaborative talk of this type, the constraints on how it is done, and what kinds of interactional purposes it serves. Extracts 6 and 7 both involve a negative assessment of a third party (a neighbour in Extract 6, another nurse in Extract 7). As in many of the cases analysed by Lerner, collaborative completion is used as a way of getting another person to articulate a negative or 32

HylandK_Chap02_Finals.indd 32

12/1/2010 8:01:11 AM

Conversation Analysis

critical assessment, thereby co-implicating them in that assessment. A speaker brings a TCU to just the point at which a co-conversationalist is able to discern (using projectability) the sort of thing they mean to say about the person, and then leaves it to them to say it. Notice that, in each case, the only thing the second speaker does is complete the first speaker’s TCU. They do not, for example, produce additional talk of their own, or respond in any way to the TCU they have completed. In other words, they treat themselves as having only what Lerner refers to as ‘conditional entry’ into the first speaker’s turn space: it is conditional in the sense that all they should properly do in that slot is complete the first speaker’s TCU on their behalf. Both participants treat the TCU as still ‘belonging’ to the first speaker: the second speaker does this by restricting what they say to completing the TCU; the first speaker does it by assessing the ‘correctness’ of the second speaker’s completion as a more or less adequate version of what they meant to say (Chris’s Yeah, Extract 6, line 6; Amy’s Yes, Extract 7, line 9). In ratifying (as in Extracts 6 and 7) or – as is also possible – repudiating (Lerner 2004) a TCU completion, the first speaker treats the TCU as, by right, their own. The capacity to complete a TCU started by a co-conversationalist is a kind of ‘mind-reading’. It demonstrates an understanding of what another speaker might be about to say, thereby offering a powerful resource for one party to an interaction to display affiliation and empathy with another. Our final data extract illustrates the use of collaborative completion in an institutional context: that of a telephone helpline. Extract 8 is drawn from a call to a helpline for women in crisis after a traumatic birth, and the caller, Hannah, is describing what happened after her emergency hysterectomy. She is cautiously building a case for the possibility that her health care team may be to blame for her hysterectomy. On the very brink of producing this negative assessment of their behaviour, Hannah slows down her talk mid-TCU (with the silence after were and the increasingly protracted sound stretches, line 9), thereby creating an opportunity for the calltaker to come in prior to completion and to complete the TCU on her behalf (line 10): Extract 8 [BCC 205] 01 Han: If things had just naturally gone wro:ng 02 .hh[hh] they’d’ve been looking at the file= 03 Clt: [Mm] 04 Han: ‘n: sayin’ “Oh this woman was just damned 05 unlucky[:”.] 06 Clt: [Y e]ah. hh 33

HylandK_Chap02_Finals.indd 33

12/1/2010 8:01:11 AM

Bloomsbury Companion to Discourse Analysis

07 08 09 10 11 12 13 14 15

Han: Uhm (0.2) But I almost felt that these:: (1.0) looking ba:ck: (0.2) that (.) these people we:re (.) coming i::n to::: Clt: cover their ba:ck[s.] Han: [c ]over themse:lves. In [some] wa:y. Clt: [Yes:] Clt: .hhh Y- Yours is a very: intelligent and reasonable response to all this:.

By speaking before Hannah’s turn is possibly complete (still in the middle of a sentential TCU and lacking a verb), and by speaking so as to voice the unspoken part of Hannah’s turn, the call-taker is engaged in an affiliative action. She is showing both that she understands Hannah’s position, and – by co-implicating herself in its articulation – that she is treating it as a reasonable position to take. Hannah’s subsequent delayed completion of her own talk with a near-repeat of what the call-taker has said (line 11) ratifies the call-taker’s completion while also reasserting authority over her turn’s talk and making relevant a next turn responsive to it. This next turn receipts Hannah’s claim and positively assesses it (lines 13–15). Displays of empathetic understanding are an important component of help-line work in ‘woman-to-woman’ organizations such as the birth crisis line, and the turn-taking model offers insights into how these are accomplished, which can be harnessed for training (Kitzinger in press).

Summary, Conclusion and Future Prospects In this chapter, we have introduced the reader to CA as a distinctive, interdisciplinary approach to understanding social life, and have briefly sketched out the six key structural features of talk-in-interaction around which its programme of enquiry is organized. Space constraints have permitted us to focus on just one of these features (the turn-taking system) and to offer just a single example of the kind of findings CA research on turn-taking is able to generate (the phenomenon of collaborative completion). We have tried, nonetheless, both to describe the key technical features of turn-taking and also to give a sense of their interactional implications for co-conversationalists as they go about the business of their everyday lives. It would, of course, be possible to undertake a similar exercise for each of the other ‘building blocks’ of social life identified by conversation analysts. CA is an immensely rich and detailed field, and its cumulative empirical findings offer us a substantial insight into the ways in which social life is lived and intersubjective understandings are constructed in and through the medium of talk-in-interaction. 34

HylandK_Chap02_Finals.indd 34

12/1/2010 8:01:11 AM

Conversation Analysis

CA is not, of course, without its critics. It is sometimes dismissed as jargonridden and impenetrable, and (despite its claims to fidelity to participants’ own orientations) as divorced from speakers’ own understandings of what is going on in interaction. For example, linguist Robin Lakoff (2003: 168–9) asks ‘who is aware that a TRP . . . is approaching?’ and ‘who realizes they are producing a dispreferred second?’ Of course most people do not use these technical terms for what they are doing in conversation, any more than they use terms like ventricles or aorta to describe the beating of their hearts or the circulation of blood around their bodies. The technical terminology is an analyst’s resource for describing what people are demonstrably doing in, for example, tracking turns at talk in order to project where a turn may be possibly complete and speaker transition relevant, and (typically) starting up just there (look again at Extracts 3 and 4). Likewise, Bud (in Extract 1) and Fran (in Extract 2) surely knew that they were producing (or beginning to produce) responses that would frustrate their co-conversationalists’ hopes of getting together any time soon – and show this through the way they have designed their turns. For a more extended discussion of CA’s commitment to participants’ orientations, see Kitzinger (2008). Other critiques depict CA as too narrow and restrictive in scope, particularly for politically engaged research (e.g. Billig 1999; Wetherell 1998), and as relying purely on snippets of decontextualized talk. One critic (reviewing theories of language and gender) claims that CA severely ‘limits admissible context’ (Bucholtz 2003: 52). In fact, CA’s notion of context is extremely broad, spanning sequential, structural, institutional and cultural aspects. The sequential meaning of context is deeply intricated into CA, and refers to prior and subsequent utterances. It is sequential context that tells us whether a short utterance (e.g. someone saying their name, as in Extract 3, line 2) – is or isn’t a possibly complete turn (here, we see it is responsive to a request for repetition). The location of an utterance within an interaction’s overall structural organization is another way of looking at its context: whether a request for antibiotics comes in the opening or diagnosis phase of a medical interaction, for example, and the consequences of its structural location, in terms of whether antibiotics are prescribed (Stivers et al. 2003). CA also examines the ways in which particular institutional contexts – for example, educational, legal, medical – are ‘talked into being’: how the talk is constitutes the context as a classroom, courtroom, or consulting room interaction, for example. And, at the broadest level, CA understands the cultural contexts in which we live as constituted by – and constitutive of – our interactions. The words we are able to select and the inferences associated with them (e.g. ‘wife’ and ‘husband’ normatively refer to opposite-sex co-resident partners, Kitzinger 2005), the grammatical, phonetic and other resources we have available to build our talk – all of these rely on, and reproduce, culture. 35

HylandK_Chap02_Finals.indd 35

12/1/2010 8:01:11 AM

Bloomsbury Companion to Discourse Analysis

Finally, despite the accumulated findings of some 50 years of CA research, there is still much we do not know about the basic structural features of interaction. ‘Basic’ CA research is continuing to expand and develop our knowledge of these basic features, building on what has gone before. A second area of research development is a detailed examination of how gesture and other body behavioural cues are integrated with talk (following Goodwin 1981 and Wootton 1997); and an understanding of embodied action as an organization of practice in its own right (work begun by Lerner and Raymond 2008). Lastly, conversation analysts are increasingly working in ‘applied’ contexts, particularly to relate conversational structures and practices to outcomes, in order to speak more directly to policy makers (e.g. Heritage et al. 2007). These are all key opportunities – and key challenges – for the future development of the field.

Transcription Key Aspects of the Relative Timing of Utterances [] =

square brackets equals sign

< ‘greater than’ sign (0.5) time in parentheses (.)

period in parentheses

overlapping talk no discernible interval between turns (also used to show that the same person continues speaking across an intervening line displaying overlapping talk ‘jump started’ talk with loud onset intervals within or between talk (measured in tenths of a second) discernable pause or gap, too short to measure

Characteristics of Speech Delivery . , ? ¿

period comma question mark inverted question mark

! :

exclamation mark hyphen/dash colon

↑↓

up or down arrow

here

underlining

closing intonation slightly upward ‘continuing’ intonation rising intonation rising intonation weaker than that indicated by a question mark animated tone abrupt cut off of sound extension of preceding sound; the more colons the greater the extension marked rise or fall in intonation immediately following the arrow emphasized relative to surrounding talk

36

HylandK_Chap02_Finals.indd 36

12/1/2010 8:01:11 AM

Conversation Analysis

HERE °here° >thisso do it now!
so do it now.ai dono wai< ↑↓ (h)

falling intonation continuing intonation exclamatory intonation Emphasis abrupt sound stop louder than surrounding talk quieter than surrounding talk Pause, timed to tenths of seconds onset of overlapping talk latched speech sound stretch questionable transcription English gloss of Pidgin (Hawai’i Creole); Pidgin transcribed using Odo orthography faster than surrounding talk rising/falling shift in intonation laugh token

Notes 1. Because this is an introductory account, we usually cite the classic, foundational studies over more recent treatments. Some of the latter can be found in the ‘Key Readings’ section below.

99

HylandK_Chap06_Finals.indd 99

12/1/2010 8:02:57 AM

Bloomsbury Companion to Discourse Analysis 2. This is true particularly of ‘applied CA’, which in contrast to ‘pure CA’ employs ‘CA concepts and methods for accomplishing its own particular [e.g. ethnographic] agenda’ (ten Have 2007: 56). 3. In this connection, Hymes (1972b) developed the notion of communicative competence – the categories of knowledge and skill needed to be a fully functioning communicator in a speech community. Communicative competence was developed in response to Chomsky’s notion of competence, which limited the scope of linguistic knowledge to grammar. For Hymes, grammatical knowledge was only one aspect of communicative competence. 4. ‘Local’ (with a capital L) is a common identity category in Hawai’i, signifying those born and raised in the islands. ‘Local ESL’, an etic category, refers to students who were institutionally identified as ESL students at Tradewinds yet displayed knowledge of and affiliation with local cultural forms and practices, including speaking Pidgin (Hawai’i Creole), the local language of Hawai’i.

Key Readings Bloome, D., Carter, S. P., Christian, B. M., Otto, S. and Shuart-Faris, N. (2005), Discourse Analysis and the Study of Classroom Language and Literacy Events: A Microethnographic Perspective. Mahwah, NJ: Erlbaum. Kamberelis, G. and Dimitriadis, G. (2005), On Qualitative Inquiry: Approaches to Language and Literacy Research. New York: Teachers College Press. Miller, G. and Fox, K. J. (2004), ‘Building bridges: the possibility of analytic dialogue between ethnography, conversation analysis and Foucault’, in D. Silverman (ed.), Qualitative Research: Theory, Method, and Practice (2nd edn). London: Sage, pp. 35–55. Saville-Troike, M. (2003), The Ethnography of Communication (3rd edn). Oxford: Blackwell. Talmy, S. (2010), ‘Critical research in applied linguistics’, in B. Paltridge and A. Phakiti (eds), Continuum Companion to Research Methods in Applied Linguistics. London: Continuum, pp. 127–42.

100

HylandK_Chap06_Finals.indd 100

12/1/2010 8:02:57 AM

7

Systemic Functional Linguistics J. R. Martin

Chapter Overview Synthesis of Current Thinking and Research Sample Study New Directions Notes Key Readings

101 104 116 119 119

Synthesis of Current Thinking and Research Systemic Functional Linguistics (hereafter SFL) is a comprehensive theory of language and social context developed principally in Britain and Australia over the past six decades. It draws on Saussure and Hjelmslev in its relational conception language as a stratified system of signs, and follows Firth in treating meaning as function in context. In addition, it provides one influential theoretical foundation for work across semiotic systems in multimodal discourse analysis (reviewed by O’Halloran this volume). Significantly SFL has evolved as an appliable linguistics (Halliday 2008a), designed to address language problems faced by the community, including educational (e.g. Martin 2009a), clinical (e.g. Fine 2006) and forensic contexts (e.g. Martin et al. 2007). Martin 2001, 2002 and 2009b provide historically contextualized reviews of discourse analysis informed by SFL, including extensive references, and can be consulted in relation to the more specific and to some extent more prospective account offered here. SFL models linguistic resources on three levels of abstraction – phonology/graphology, (realizing) lexicogrammar, (realizing) discourse semantics. Higher strata involve emergently complex patterns of lower strata ones; all levels make meaning. In addition, resources on each stratum are organized metafunctionally, according to the kind of meaning they construe – that is, ideational resources naturalizing physical/biological materiality and semiosis, 101

HylandK_Chap07_Finals.indd 101

12/1/2010 8:03:17 AM

Bloomsbury Companion to Discourse Analysis

interpersonal resources negotiating social relations and textual resources managing information flow. This hierarchy of realization is outlined in relation to its attendant complementarity of metafunction in Figure 7.1. Following Firth and Hjelmslev, SFL models social context as more abstract levels of semiosis; the level next to language is mapped metafunctionally as field (ideational context), tenor (interpersonal context) and mode (textual context). Field is concerned with social activity, across all walks of life – including home, recreation, trades and crafts, professions and disciplines. Tenor is concerned with social relations, negotiated in relation to power and solidarity. Mode is concerned with the effect of various technologies of communication on the texture of information flow – speaking vs writing for example, alongside various electronic modalities (radio, TV, telephone, texting, email, blogging etc.). These three register variables are coordinated in relation to social purpose by the higher level of genre, which specifies which field, tenor and mode variables map onto one another in a given culture and how their realization is staged in phases of unfolding discourse. This social semiotic model of language and context is summarized in Figure 7.2. Stratification and metafunction provide the theoretical parameters for the descriptive cartography deployed by SFL text analysts. English resources are mapped along these lines in Table 7.1, including key references to the descriptions commonly deployed (inclusive of Kress and Van Leeuwen’s work on images which is so often deployed in inter-modal verbiage/image analysis). As exemplified in Martin 2009c, the basic rule of thumb in using this matrix is to shunt among cells, taking care to look upwards to higher levels for contextualization and rightwards to textual meaning for co-textualization.

discourse semantics textual interpersonal

lexicogrammar ideational phonology

Figure 7.1 Basic SFL parameters – stratification and metafunction

102

HylandK_Chap07_Finals.indd 102

12/1/2010 8:03:18 AM

Systemic Functional Linguistics

genre tenor mode interpersonal textual

field ideational

Figure 7.2 Language strata in relation to social context (stratified as register and genre)

Table 7.1 Metafunction/stratum matrix of English resources for text analysis metafunction stratum

ideational

interpersonal

textual

genre Martin and Rose 2008 Eggins and Slade 1997 register Martin 1992 Halliday and Martin 1993 Christie and Martin 1997 Martin and Veel 1998 Martin and Wodak 2003 Christie and Martin 2007 discourse semantics Martin and Rose 2003/2007 Martin and White 2005

orbital/serial structure

prosodic structure

periodic structure

field – activity sequences, participant taxonomies

tenor – power, solidarity

mode – action/ reflection; monologue/ dialogue

ideation, external conjunction

appraisal, negotiation

identification, internal conjunction, information flow (Continued)

103

HylandK_Chap07_Finals.indd 103

12/1/2010 8:03:18 AM

Bloomsbury Companion to Discourse Analysis

Table 7.1

(Cont’d)

metafunction stratum

ideational

interpersonal

textual

lexicogrammar [verbiage] Halliday and Matthiessen 2004 [image] Kress and van Leeuwen 1996/2006

transitivity; nominal group classification, description, enumeration

mood, modality, polarity, comment, vocation; nominal group attitude, person

graphology/phonology Halliday and Greaves 2008

tone sequence

formatting, emoticons, colour; tone, voice quality, phonaesthesia

theme and information; tense and deixis; ellipsis and substitution punctuation, layout; tonality, tonicity

Figure 7.3 GW Living Modern Guru column (15 November 2008)

Sample Study By way of illustration we will consider following text, ‘Testing times’, which appeared in the GW Living section of the Sydney Morning Herald’s weekly Good Weekend magazine, on 15 November 2008 (Katz 2008) – especially its opening. It is from a regular ‘agony aunt’ column entitled ‘Modern Guru’ in which 104

HylandK_Chap07_Finals.indd 104

12/1/2010 8:03:18 AM

Systemic Functional Linguistics

Danny Katz spoofs the genre, whimsically responding to readers’ questions about ‘21st-century ethics, etiquette and dilemmas’ (column reproduced as Figure 7.3). Testing Times: how to handle one mark too many I am 16 years old and still at school. Recently I got an exam back and the teacher had added up my scores incorrectly and given me an extra mark. Should I have told her about the mistake, or just kept the extra mark? E.B., Pymble, NSW Ask just about any schoolkid and here’s what they’d probably say: ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ – and, by the way, this is how all schoolkids talk; I know, I’ve snuck peeks at my own kids’ msn messages, until they blocked me out using the parental control filter. Apparently it can also filter out controlling parents. But just ask about any schoolteacher and here’s what they’d probably say: ‘You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting in a botched education, culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ – and by the way, that’s how I imagine all school teachers talk, and they all wear mortarboards and black gowns, and look like old wire-moustached Latin masters as drawn by Ronald Searle. But ask the rest of us, and we’d probably go a bit both ways, because sometimes life throws you lovely little windfalls that you should be able to enjoy without guilt – that extra mark on an exam paper, that accidental $10 from a faulty ATM, that unexpected meatball in your turkey-breast sub, these are some of the great moments of life and should be cherished. But at the same time, there are rules when it comes to windfalling: your lucky little bonus must be small enough that nobody gets hurt, small enough so you can enjoy it without a heavy conscience, and most importantly, small enough so you can feign ignorance if you get caught. Which is why I propose a Universal Fortuitous Windfall CutOff Point before a person needs to advise the relevant authorities – and that cut-off point is three. You’re allowed up to three extra marks on an exam paper before telling the teacher. Up to three $10 notes from a faulty ATM before returning the cash to the bank. And up to three unasked-for meatballs on a non-meatball sub before you yell at the Subway girl. The advice column opens as follows, presented here divided into ranking1 clauses (i.e. non-embedded ones), with clause ellipses of the kind specified in 105

HylandK_Chap07_Finals.indd 105

12/1/2010 8:03:20 AM

Bloomsbury Companion to Discourse Analysis

Halliday and Hasan’s 1976 work on cohesion filled in and placed in parentheses below. I’ve put a line break between orthographic sentences, and double square brackets around the one embedded clause. [1] [2] [3] [4] [5]

I am 16 years old and (I am) still at school. Recently I got an exam back and the teacher had added up my scores incorrectly and (she had) given me an extra mark.

[6] Should I have told her about the mistake, [7] or (should I have) just kept the extra mark? [8] [9] [10] [11] [12] [13] [14] [15]

Ask just about any schoolkid and here’s [[what they’d probably say]]: ‘Nawwwww (you shouldn’t have told her), keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ –

With reference to Table 7.1, we’ll focus on discourse semantics. In Working with Discourse Martin and Rose (2007 edition) outline five key systems alongside their realization in relation to information flow (periodicity). We will explore these, in more and less detail, in the following order here: – identification – conjunction – ideation – negotiation – appraisal Identification is concerned with discourse entities – with people, places and things, and the way they are introduced in a text and kept track of once there. As compiled in Table 7.2, our sample text involves four entities that are anaphorically interdependent, thus forming cohesive reference chains. Pronouns, a proper name, phoric determiners (the, that) and ellipsis (noted in parentheses below) are used to keep track of participants: the student E.B., his/her teacher, the extra mark and schoolkids in general. In addition, the text makes use of extended reference (Halliday and Hasan 1976) to identify larger configurations of meaning with one another. As presented in Table 7.3, this involves two instances of anaphoric reference (the mistake, that) and one of cataphora (here).

106

HylandK_Chap07_Finals.indd 106

12/1/2010 8:03:20 AM

Systemic Functional Linguistics

Table 7.2 Identification chains E.B.

teacher

mark

schoolkid

I (I) I my me I (I) you E.B.

the teacher (she) her (her) the teacher she

an extra mark the extra mark that extra mark

any schoolkid they

Table 7.3 Extended reference the teacher had added up my ⇐ the mistake scores incorrectly and (the teacher had) given me an extra mark. ‘Nawwwww (you shouldn’t have told her), keep that extra here ⇒ mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ specially if it was, like, a maths ⇐ that test and she added up wrong

Conjunction focuses on logical relations of addition, comparison, time and cause between figures (i.e. configurations of process, participant and circumstance, after Halliday and Mattheissen 1999). A reticulum displaying these relations is presented in Figure 7.4. By convention, internal relations construing the rhetorical organization of the text are modelled on the left hand side of the diagram, and external relations construing relations of comparison, time and cause among ‘real world’ events on the right (external additive relations are modelled down the centre of the reticulum). The actual realization of explicit conjunctive relations in the text, by conjunctions (or tense) is noted in italics; items potentially realizing implicit relations are placed in parentheses. Lexicogrammatically, conjunctive relations have to be realized in either the first or second of related units; when realized in the first unit dependency arrows point forward (e.g. if in unit 13 above), and when realized in the second, arrows point back, indicating retrospective dependency. As far as external relations are concerned, Figure 7.4 outlines the complicated sequencing involved in telling a story out of time, since the events in units 4 and 5 precede that in unit 3, which is in turn followed by the alternatives in units 6

107

HylandK_Chap07_Finals.indd 107

12/1/2010 8:03:20 AM

Bloomsbury Companion to Discourse Analysis 1 exp/add

and

2 3 exp/add

exp/prec

and, had added

4 exp/add

imp/succ

and, (then)

imp/succ

(then)

5 6 exp/alt

or

7 8 exp/add

imp/cond

and, (then)

exp/cond

if

9 (namely)

imp/i.e.

(like)

imp/i.e.

coz

exp/caus

specially

exp/i.e

10 11 12 13 exp/add 14 15

Figure 7.4 Conjunctive relations (imp = implicit, exp = explicit; i.e. = reformulation, caus = causal, cond = conditional, add = additive, alt = alternative, succ = succeeding, pred = preceding)

and 7. Internally, Figure 7.4 displays the interplay of motivation and reformulation involved in argumentation, with unit 9 exemplified by units 10–15, unit 10 elaborated as unit 11–15, unit 11 justified by units 12–15 and unit 12 specified by units 13 and 14. As can be seen, reticula treat texts as simultaneously structured by internal and external relations, and as unfolding serially from one interdependent unit to another. For detailed discussion of this kind of representation in relation to Rhetorical Structure Theory (RST), which favours a nucleus and satellite analysis, generally privileging one unit as central (unit 11 perhaps above), see Martin 1992. Unlike conjunctive relations, RST requires that every unit be related to 108

HylandK_Chap07_Finals.indd 108

12/1/2010 8:03:20 AM

Systemic Functional Linguistics

another, including relations that SFL would model as negotiation or genre; this means that unlike conjunctive relations, RST is not restricted to relations that can potentially be realized by a conjunction. For the perspective of ideation, the external temporal relations reflect two short episodes of sequencing. Units 3–7 sequence the figures in which the teacher adds up the scores incorrectly, gives the extra mark, returns the exam and the student either alerts the teacher or keeps the mark. teacher add up score incorrectly ^teacher give schoolkid extra mark ^schoolkid get exam back ^schoolkid tell teacher about mark/student keep extra mark And units 8 and 9 sequence the turns of the imagined dialogue: (someone) ask schoolkid ^schoolkid say (to someone) As far as nuclear relations within figures are concerned, the teacher’s error is configured five times, including one lexicalization (mistake) and one instance of extended reference (that); in the examples below, ⫹ marks the extension of the process by a medium, ⫻ its enhancement by a circumstance, and ⫻⫹ its extension/enhancement by an agent or beneficiary. teacher ⫻⫹ add up ⫹ scores ⫻ incorrectly mistake teacher ⫹ stuff up teacher ⫹ add up ⫻ wrong [teacher ⫹ add up ⫻ wrong] Telling the teacher about the mark is realized twice (once via the elliptical response Nawwwww), as is keeping the mark. schoolkid ⫹ tell ⫻⫹ teacher ⫻ about mark (schoolkid ⫹ tell ⫻⫹ teacher ⫻ about mark) schoolkid ⫻⫹ keep ⫹ mark (schoolkid) ⫻⫹ keep ⫹ mark Turning to entities, the ideational focus of these sequences and figures is reinforced through repeated realizations of the student, teacher and exam. The schoolkid and teacher strings are constructed through repetition (teacherteacher etc.) and an instance of meronymy (school-schoolkid). The exam string 109

HylandK_Chap07_Finals.indd 109

12/1/2010 8:03:21 AM

Bloomsbury Companion to Discourse Analysis

is more varied, including repetition (mark-mark), synonymy (score-mark), meronymy (exam-mark) and hyponymy (exam-maths test). These strings are presented below, including lexical rendering of all pronominal reference (in square brackets) and ellipses (in parentheses). school, [schoolkid], (schoolkid), [schoolkid], [schoolkid], [schoolkid], [schoolkid], (schoolkid), schoolkid, [shoolkid], (schoolkid), (schoolkid) teacher, (teacher), [teacher], [teacher], teacher, [teacher] exam, score, mark, mark, mark, maths test Overall the activity sequencing, and nuclear and taxonomic relations clearly position the text in the field of education – with respect to what is going on and the participants involved. Turning to interpersonal meaning, the text involves a double barrelled question followed by an imagined peer response. The response first addresses the first part of the question (unit 6), which it treats as seeking information, responding in the negative: K2 K1

[6] Should I have told her about the mistake, - [10] ‘Nawwwww (you shouldn’t have told her),

It then responds to unit 7 as if it was an offer rather than a question, advising E.B. to keep the mark, and then justifying that advice. The Modern Guru does not make room for a putative response from E.B. since he is about to consider an alternative response from an imaginary teacher and then proffer one of his own. Da1 [7] or (should I have) just kept the extra mark? A2 - [11] keep that extra mark justify [12] coz, like, the teacher totally stuffed up, y’know, [13] specially if it was, like, a maths test [14] and she added up wrong, haw haw, [15] that’d be, like hilarical’ – The first response negotiates E.B.’s query as concerned with exchanging knowledge, notated as a K2-K1 sequence (where K2 stands for a secondary knower move, seeking information, and K1 for a primary knower move, providing it). The second response renegotiates the query as concerned with action, notated as Da1-A2-justify (where Da1 stands for a delayed move by the primary actor, checking whether to act, A2 for a secondary actor move, telling the primary actor what to do, followed up by rationale for the advice given). For a detailed account of this style of conversation analysis see Martin 1992; Martin and Rose 2003 (2007 edition); Ventola 1987. The complementary responses capture nicely 110

HylandK_Chap07_Finals.indd 110

12/1/2010 8:03:21 AM

Systemic Functional Linguistics

the dual nature of asking for and giving advice – since this involves exchanging information (knowledge) about what to do (action). Turning to appraisal, we can determine the interpersonal motivation for the peer advice given. In justifying his/her counsel the imagined schoolkid judges the teacher’s behaviour negatively as incompetent (stuffed up, wrong), picking up on E.B.’s negative judgements (incorrectly, mistake), delights affectually in the supposition that she might even have been a math teacher (haw haw) and then appreciates the refined scenario as a hilariously farcical (hilarical). The basic rhetoric of feeling here, from a student perspective, is that teacher’ error is something to savour happily – and intensely, as the amplified grading highlights (Nawwwww, totally stuffed up, haw haw, hilarical). For details on the analysis of evaluation touched on here see Martin and White 2005 (judgement in bold, affect in italics and appreciation underlined below). [1] [2] [3] [4] [5]

I am 16 years old and (I am) still at school. Recently I got an exam back and the teacher had added up my scores incorrectly and (the teacher had) given me an extra mark.

[6] Should I have told her about the mistake, [7] or (should I have) just kept the extra mark? [8] [9] [10] [11] [12] [13] [14] [15]

Ask just about any schoolkid and here’s [[what they’d probably say]]: ‘Nawwwww (you shouldn’t have told her), keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical –

Although for reasons of space we will not be pursuing a periodicity analysis here (in spite of the ‘look right in Table 7.3’ prescription given above!), note that all of the inscribed evaluation in fact comes last in the clause, in the unmarked position for tonic prominence signalling News.2 It thus contrasts with topical Theme selections, boxed3 above, which are realized at the other end of the English clause and here establish schoolkids and the teacher as the text’s orientation to its education field. As we can see, unlike say conversation analysis (CA) or RST, the approach to discourse analysis exemplified here is a modular one which disperses the description across a range of metafunctionally complementary discourse systems and structures. Each of these systems is realized lexicogrammatically (which stratum 111

HylandK_Chap07_Finals.indd 111

12/1/2010 8:03:21 AM

Bloomsbury Companion to Discourse Analysis

is in turn realized phonologically or graphologically); and each of them is realizing social context (the systems of register and genre). Space precludes consideration of the lower levels here (see, however, Martin 2010 for a glimpse of their contribution to the ‘Testing Times’ text). Ultimately it is the special responsibility of the highest level of abstraction in the model, genre, to coordinate the meanings arising from different strata and metafunctions. As we can also see, the analysis is a painstaking one, ideally involving a very close reading of every meaning a text construes, on the assumption that nothing is there by accident and so has to be understood as part of the social function of the text as a whole. What about ‘looking up’ to social context, in what little space remains. Beginning then with register, field it will be recalled is concerned with institutional activity – our participation in domestic, recreational, devotional, governmental and professional life. This Modern Guru column deals principally with education, and so the rich co-patterning of lexis dealing with participants and processes in that field we have been examining extends throughout the ‘Testing Times’ text (lexicalizations underlined below). Testing Times: how to handle one mark too many I am 16 years old and still at school. Recently I got an exam back and the teacher had added up my scores incorrectly and given me an extra mark. Should I have told her about the mistake, or just kept the extra mark? . . . Ask just about any schoolkid and here’s what they’d probably say: ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ – and, by the way, this is how all schoolkids talk; . . . But just ask about any schoolteacher and here’s what they’d probably say: “You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting in a botched education,... – and by the way, that’s how I imagine all schoolteachers talk, and they all wear mortarboards and black gowns, and look like old wire-moustached Latin masters as drawn by Ronald Searle. . . . – that extra mark on an exam paper, . . . You’re allowed up to three extra marks on an exam paper before telling the teacher. Four other fields are more sketchily constituted: electronic communication, the nineteenth-century British workhouse, contemporary banking and fast food. . . . I know, I’ve snuck peeks at my own kids’ msn messages, until they blocked me out using the parental control filter. Apparently it can also filter out controlling parents. 112

HylandK_Chap07_Finals.indd 112

12/1/2010 8:03:21 AM

Systemic Functional Linguistics

culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ – . . . that accidental $10 from a faulty ATM, . . . Up to three $10 notes from a faulty ATM before returning the cash to the bank. . . . . . . that unexpected meatball in your turkey-breast sub, . . . And up to three unasked-for meatballs on a non-meatball sub before you yell at the Subway girl. Tenor is concerned with social positioning – our status in relation to one another, and our degree of affinity (power and solidarity as they are typically termed). These social relations are strongly foregrounded in the contrast the Guru sets up between how schoolkids talk to one another (equal status, close friends) and how teachers talk to them (unequal status, collegial contact). Ask just about any schoolkid and here’s what they’d probably say: ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ – and, by the way, this is how all schoolkids talk But just ask about any schoolteacher and here’s what they’d probably say: ‘You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting in a botched education, culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ – and by the way, that’s how I imagine all schoolteachers talk. The quoted speech in these examples also strongly implicates the third register variable mode, since schoolkids talk spoken English and schoolteachers talk like books (Halliday 2008b). So kids mean what they say and say what they mean, with processes realized verbally (keep, stuffed up, added up), qualities realized adjectivally (hilarical) and logical connections realized conjunctively (coz, if, and); for teachers the relation between meaning and wording is less direct (Zhu 2008), with processes and qualities regularly textured nominally (abilities, failures, destitution) and logical relations rendered verbally (lead, resulting, culminating). ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ ‘You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting 113

HylandK_Chap07_Finals.indd 113

12/1/2010 8:03:21 AM

Bloomsbury Companion to Discourse Analysis

in a botched education, culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ The final step we need to take as far as realization is concerned is to further contextualize our Testing Times text in relation to genre. Readers will no doubt have noticed during our discussion of field above that one dimension of the institutional activities implicated in the Modern Guru text has been postponed – namely the ‘agony aunt’ discourse reflected in the text’s concern with ethics and etiquette (lexicalizations underlined below), including it’s questionanswer format and modulated proposals.

Modern Guru Danny Katz answers readers’ questions about 21st-century ethics, etiquette and dilemmas. Testing Times: how to handle one mark too many . . . Should I have told her about the mistake, or just kept the extra mark? . . . But at the same time, there are rules when it comes to windfalling: your lucky little bonus must be small enough that nobody gets hurt, small enough so you can enjoy it without a heavy conscience, and most importantly, small enough so you can feign ignorance if you get caught. Which is why I propose a Universal Fortuitous Windfall Cut-Off Point before a person needs to advise the relevant authorities – and that cut-off point is three. You’re allowed up to three extra marks on an exam paper before telling the teacher. Up to three $10 notes from a faulty ATM before returning the cash to the bank. And up to three unasked-for meatballs on a non-meatball sub before you yell at the Subway girl. This dimension of field is unlike the others in that it is more fully implicated in the organization of the text as a whole. We are dealing with the consequences of writing a test in school here, not those of sending msn messages, getting money out of an ATM or ordering food at Subway – and we are offering advice in general about how to behave in relation to fortuitous windfalls. This social purpose in effect transcends the others, and in doing so affects all metafunctions, coordinating them as a staged goal-oriented social process that we recognize as genre. This genre has been explored by Thibault (1986) and will not be discussed in detail here. In terms of staging, the text has a Question/problem followed by Answer/solution structure. The Answer/solution can be divided into Rationale and Advice. This Rationale explores one extreme position on the issue, then another, before coming down in between; it is doubtful this phasing strategy can be generalized to the genre as a whole and so will not be 114

HylandK_Chap07_Finals.indd 114

12/1/2010 8:03:22 AM

Systemic Functional Linguistics

labelled here (for discussion of stages and phases in genre analysis see Martin and Rose 2008).

Question/problem I am 16 years old and still at school. Recently I got an exam back and the teacher had added up my scores incorrectly and given me an extra mark. Should I have told her about the mistake, or just kept the extra mark?

Answer/solution Rationale Ask just about any schoolkid and here’s what they’d probably say: ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ – and, by the way, this is how all schoolkids talk; I know, I’ve snuck peeks at my own kids’ msn messages, until they blocked me out using the parental control filter. Apparently it can also filter out controlling parents. But just ask about any schoolteacher and here’s what they’d probably say: ‘You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting in a botched education, culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ – and by the way, that’s how I imagine all school teachers talk, and they all wear mortarboards and black gowns, and look like old wire-moustached Latin masters as drawn by Ronald Searle. But ask the rest of us, and we’d probably go a bit both ways, because sometimes life throws you lovely little windfalls that you should be able to enjoy without guilt – that extra mark on an exam paper, that accidental $10 from a faulty ATM, that unexpected meatball in your turkeybreast sub, these are some of the great moments of life and should be cherished. But at the same time, there are rules when it comes to windfalling: your lucky little bonus must be small enough that nobody gets hurt, 115

HylandK_Chap07_Finals.indd 115

12/1/2010 8:03:22 AM

Bloomsbury Companion to Discourse Analysis

small enough so you can enjoy it without a heavy conscience, and most importantly, small enough so you can feign ignorance if you get caught.

Advice Which is why I propose a Universal Fortuitous Windfall Cut-Off Point before a person needs to advise the relevant authorities – and that cut-off point is three. You’re allowed up to three extra marks on an exam paper before telling the teacher. Up to three $10 notes from a faulty ATM before returning the cash to the bank. And up to three unasked-for meatballs on a non-meatball sub before you yell at the Subway girl.

New Directions The best resources for exploring new directions in discourse analysis in SFL is Bednarek and Martin 2010 (see also Hasan et al. 2005a, b; Halliday and Webster 2009). This edited collection focuses on two relatively unexplored hierarchies in SFL, instantiation and individuation, in relation to genesis. Instantiation is the cline relating system and text. Unlike realization, which is a hierarchy of abstraction, instantiation is a hierarchy of generality.

system

genre/ register)

text type text reading

Figure 7.5 The hierarchy of instantiation – sub-potentialization in relation to system use

116

HylandK_Chap07_Finals.indd 116

12/1/2010 8:03:22 AM

Systemic Functional Linguistics

Instantiation relates a culture’s systems of meanings as a whole to their specialization as registers and genres; at the same time, it generalizes recurring patterns of meaning across instances as text types4; and from the perspective of critical theory, texts themselves can be interpreted as potentials since they afford readings of different kinds according to the social subjectivity of their consumers. Overall, what we are looking at is a scale of potentiality – all of the meanings a semiotic system allows in relation to their sub-potentialization as instances of language use (a cricketer’s batting average, in other words, in relation to his shots and the way an umpire has adjudicated them as runs or not). A crude outline of this scale is presented as Figure 7.5. Whereas instantiation refers to the specialization of the meaning potential of a culture text by text, individuation specializes that meaning potential according to people (how meaning is deployed in relation to users rather than uses of language). The Modern Guru foregrounds individuation in the contrast he sets up between the schoolkid’s and schoolteacher’s response to the extra mark: ‘Nawwwww, keep that extra mark coz, like, the teacher totally stuffed up, y’know, specially if it was, like, a maths test and she added up wrong, haw haw, that’d be, like hilarical’ ‘You caddish little rotter! By pilfering that extra point, not only are you behaving reprehensibly, but you are failing to recognise your true scholastic abilities, which could lead to ongoing exam failures, resulting in a botched education, culminating in a life of destitution at a Dickensian workhouse, blacking boots for Mr Bumble!’ And he explicitly claims these responses to be representative of the way all schoolkids and all schoolteachers talk: – and, by the way, this is how all schoolkids talk; I know, I’ve snuck peeks at my own kids’ msn messages, until they blocked me out using the parental control filter. Apparently it can also filter out controlling parents. – and by the way, that’s how I imagine all schoolteachers talk, and they all wear mortarboards and black gowns, and look like old wire-moustached Latin masters as drawn by Ronald Searle. In SFL the main work on individuation has been oriented to generation, gender and class in the language of the mothers of pre-school children (Hasan 2009). For further discussion of the user-oriented hierarchy presented in Figure 7.6 see Martin 2009d. 117

HylandK_Chap07_Finals.indd 117

12/1/2010 8:03:22 AM

Bloomsbury Companion to Discourse Analysis

culture master identity

affiliation

sub-culture

individuation

persona

Figure 7.6 Individuation and affiliation

system/ reservoir

instantiation text logogenesis

phylogenesis repertoire ontogenesis

Figure 7.7 Realization, instantiation and individuation in relation to genesis

Current research on instantiation and individuation has rekindled interest in genesis, and the issue of how to model time as a variable in relation to meaning in unfolding discourse (logogenesis), individual development (ontogenesis) and cultural evolution (phylogenesis). The relation of these complementary perspectives on time in relation to realization, instantiation and individuation is outlined in Figure 7. 7. 118

HylandK_Chap07_Finals.indd 118

12/1/2010 8:03:23 AM

Systemic Functional Linguistics

Notes 1. Unless otherwise specified the grammar analysis deployed here is taken from Halliday and Matthiessen 2004. 2. To these we might add three instances of the bonus mark, which is graded (extra) to flag appreciation. 3. To these we might arguably add the implied ‘schoolkid’ Subjects of imperative clauses 8 and 11, and arguably subtract it in 13 by taking the dependent conditional clause if it was, like, a maths test as a marked Theme. 4. Halliday and Matthiessen (e.g. 1999, 2004) do not distinguish between genre/register and text type as different levels on this scale, treat text as the instance end of the cline, and do not of course explicitly position genre on the scale (since they do not operate with a stratified model of context).

Key Readings Halliday and Greaves outline English resources for intonation; Halliday and Matthiessen is the foundational text for English grammatical resources, extended to other language families in Caffarel et al. Martin and Rose 2003/2007 is an accessible introduction to discourse semantics, elaborated for evaluative language in Martin and White, and applied to casual conversation in Eggins and Slade. Christie and Martin comprises studies of school and workplace registers and genres, and Martin and Rose 2008 provides a basic introduction to SFL genre theory. For reasons of space I have foregrounded my own SFL perspective on discourse analysis here; for a broader purchase see Halliday and Webster 2009. Caffarel, A., Martin, J. R. and Matthiessen, C. M. I. M. (eds) (2004), Language Typology: A Functional Perspective. Amsterdam: Benjamins. Christie, F. and Martin, J. R. (eds) (1997), Genre and Institutions: Social Processes in the Workplace and School. London: Pinter (Open Linguistics Series). Eggins, S. and Slade, D. (1997), Analysing Casual Conversation. London: Cassell [reprinted Equinox 2005]. Halliday, M. A. K. and Greaves, W. S. (2008), Intonation in the Grammar of English. London: Equinox. Halliday, M. A. K. and Matthiessen, C. M. I. M. (2004), An Introduction to Functional Grammar (3rd edn). London: Arnold. Halliday, M. A. K. and Webster, J. (eds) (2009), Continuum Companion to Systemic Functional Linguistics. London: Continuum. Martin, J. R. and Rose, D. (2003), Working with Discourse: Meaning beyond the Clause. London: Continuum [2nd revised edn 2007]. Martin, J. R. and Rose, D. (2008), Genre Relations: Mapping Culture. London: Equinox. Martin, J. R. and White, P. R. R. (2005), The Language of Evaluation: Appraisal in English. London: Palgrave.

119

HylandK_Chap07_Finals.indd 119

12/1/2010 8:03:23 AM

8

Multimodal Discourse Analysis Kay L. O’Halloran

Chapter Overview Introduction Approaches to MDA Theoretical and Analytical Issues in MDA Sample MDA Text Analysis New Directions in MDA Notes Key Readings

120 122 124 127 136 136 136

Indeed, we can define a culture as a set of semiotic systems, a set of systems of meaning, all of which interrelate. (Halliday and Hasan 1985: 4)

Introduction Multimodal discourse analysis (henceforth MDA) is an emerging paradigm in discourse studies which extends the study of language per se to the study of language in combination with other resources, such as images, scientific symbolism, gesture, action, music and sound. The terminology in MDA is used somewhat loosely at present as concepts and approaches evolve in this relatively new field of study. For example, language and other resources which integrate to create meaning in ‘multimodal’ (or ‘multisemiotic’) phenomena (e.g. print materials, videos, websites, three-dimensional objects and day-to-day events) are variously called ‘semiotic resources’, ‘modes’ and ‘modalities’. MDA itself is referred to as ‘multimodality’, ‘multimodal analysis’, ‘multimodal semiotics’ and ‘multimodal studies’.

120

HylandK_Chap08_Finals.indd 120

12/1/2010 12:58:12 PM

Multimodal Discourse Analysis

For the purpose of clarity, in this chapter semiotic resource is used to describe the resources (or modes) (e.g. language, image, music, gesture and architecture), which integrate across sensory modalities (e.g. visual, auditory, tactile, olfactory, gustatory, kinesthetic) in multimodal texts, discourses and events, collectively called multimodal phenomena. Following Halliday (1978: 123), semiotic resources are ‘system[s] of meanings that constitute “the ‘reality” of the culture’. The medium is the means through which the multimodal phenomena materialize (e.g. newspaper, television, computer or material object and event). In what follows, the major concerns of MDA, the reasons for the emergence of this field in linguistics, and the variety of approaches which have been developed are discussed, before concepts specific to MDA are examined in more detail and a sample multimodal analysis is presented. MDA is concerned with theory and analysis of semiotic resources and the semantic expansions that occur as semiotic choices combine in multimodal phenomena. The ‘inter-semiotic’ (or inter-modal) relations arising from the interaction of semiotic choices, known as intersemiosis, is a central area of multimodal research (Jewitt 2009a). MDA is also concerned with the design, production and distribution of multimodal resources in social settings (e.g. van Leeuwen 2008), and the resemioticization (Iedema 2001b, 2003) of multimodal phenomena which takes place as social practices unfold. The major challenges facing MDA include the development of theories and frameworks for semiotic resources other than language, the modelling of social semiotic processes (in particular, intersemiosis and resemioticization), and the interpretation of the complex semantic space which unfolds within and across multimodal phenomena. There are several reasons for the paradigmatic shift away from the study of language alone to the study of the integration of language with other resources. First, discourse analysts attempting to interpret the wide range of human discourse practices have found the need to account for the meaning arising from multiple semiotic resources deployed in various media, including contemporary interactive digital technologies. Second, technologies to develop new methodological approaches for MDA, for example multimodal annotation tools (Rohlfing et al. 2006) have become available and affordable. Lastly, interdisciplinary research has become more common as scientists from various disciplines seek to solve similar problems. From ‘an age of disciplines, each having its own domain, its own concept of theory, and its own body of method’, the twentieth century has emerged as ‘age of themes’ (Halliday 1991: 39) aimed at solving particular problems. MDA is an example of this paradigm shift, and it has a key contribution to make with respect to multimodal analysis, search and retrieval of information.

121

HylandK_Chap08_Finals.indd 121

12/1/2010 12:58:13 PM

Bloomsbury Companion to Discourse Analysis

Approaches to MDA Gunther Kress and Theo van Leeuwen (1996 [2006]) and Michael O’Toole (1994, 2010) provided the foundations for multimodal research in the 1980s and 1990s, drawing upon Michael Halliday’s (1978, 1985 [1994, 2004]) social semiotic approach to language to model the meaning potential of words, sounds and images as sets of interrelated systems and structures. Kress and van Leewuen (2006) explored images and visual design, and O’Toole (2010) applied Halliday’s systemic functional model to a semiotic analysis of displayed art, paintings, sculpture and architecture. Halliday’s (1978; Halliday and Hasan, 1985) concern with both text and context, instance and potential, is reflected in these foundational works. That is, Kress and van Leeuwen (2006) adopt a (top-down) contextual approach with a particular orientation to ideology, deriving general principles of visual design which are illustrated via text analysis; while O’Toole (2010) develops a (bottom-up) grammatical approach by working closely with specific ‘texts’ (i.e. paintings, architectural designs and sculptures) to derive frameworks which can be applied to other works. Subsequent research has built upon these two approaches and extended them into new domains. For example, contextual approaches have been developed for speech, sound and music (van Leeuwen 1999), scientific texts (Lemke 1998), hypermedia (Lemke 2002), action and gesture (Martinec 2000), educational research (Jewitt 2006) and literacy (Kress 2003). In addition, grammatical approaches to mathematics (O’Halloran 2005), hypermedia (Djonov 2007) and a range of other multimodal texts (e.g. Bednarek and Martin, 2010) have resulted in an approach which has been called systemicfunctional multimodal discourse analysis (SF-MDA). Jewitt (2009b: 29–33) classifies contextual and grammatical approaches as ‘social semiotic multimodality’ and ‘multimodal discourse analysis’ respectively. These approaches provide complementary perspectives, being derived from Michael Halliday’s social semiotic approach to text, society and culture (see Iedema 2003), which grounds social critique in concrete social practices through three fundamental principles: (1) Tri-stratal conceptualization of meaning which relates low level features in the text (e.g. images and sound) to higher-order semantics through sets of interrelated lexicogrammatical systems, and ultimately to social contexts of situation and culture. (2) Metafunctional theory which models the meaning potential of semiotic resources into three distinct ‘metafunctions’: • Ideational meaning (i.e. our ideas about the world) involves: − Experiential meaning: representation and portrayal of experience in the world. 122

HylandK_Chap08_Finals.indd 122

12/1/2010 12:58:13 PM

Multimodal Discourse Analysis

− Logical meaning: construction of logical relations in that world. • Interpersonal meaning: enactment of social relations. • Textual meaning: organization of the meaning as coherent texts and

units. (3) Instantiation models the relations of actual choices in text to the systemic potential, with intermediate sub-potentials – registers – appearing as patterns of choice in text-types (e.g. casual conversation, debate and scientific paper). Multimodal research rapidly expanded in mid-2000s onwards as systemic linguists and other language researchers became increasingly interested in exploring the integration of language with other resources. There was an explicit acknowledgement that communication is inherently multimodal and that literacy is not confined to language. Further approaches to multimodal studies evolved. These include Ron Scollon, Suzanne Wong Scollon and Sigrid Norris’ multimodal interactional analysis (Norris 2004; Norris and Jones 2005; Scollon 2001; Scollon and Wong Scollon 2004), developed from mediated discourse analysis which has foundations in interactional sociolinguistics and intercultural communication, and Charles Forceville’s (Forceville and Urios-Aparisi 2009) cognitive approach to multimodal metaphor based on cognitive linguistics (Lakoff and Johnson 1980). In addition, critical discourse approaches have been developed (Machin 2007; van Leeuwen 2008), based on social semiotics and other critical traditions. A variety of distinct theoretical concepts and frameworks continue to emerge in multimodal studies (see Jewitt 2009c), but most have some relationship to one or more of these paradigms. The increasing popularity of MDA is evidenced by recent publications (e.g. Baldry and Thibault 2006; Bateman 2008; Bednarek and Martin 2010; Jewitt 2009c; Unsworth 2008; Ventola and Moya 2009). Unsurprisingly, there is much debate about the nature of this emerging field (Jewitt 2009c). While multimodality can be characterized as ‘a domain of enquiry’ (Kress 2009: 54) (e.g. visual design, displayed art, mathematics, hypermedia, education and so forth), theories, descriptions and methodologies specific to MDA are clearly required (O’Halloran and Smith, in press) and some frameworks and tools have indeed already been developed (e.g. Bateman 2008; Bednarek and Martin 2010; Lemke 2009; O’Halloran 2005; O’Toole 2010). As a domain of enquiry, multimodal studies encourage engagement and cross-fertilization with other disciplines which have the same object of study. Incorporating knowledge, theories and methodologies from other disciplines poses many problems, however, not least being the provision of adequate resources for research to be undertaken across traditional disciplinary boundaries. 123

HylandK_Chap08_Finals.indd 123

12/1/2010 12:58:13 PM

Bloomsbury Companion to Discourse Analysis

The development of theories and practices specific to MDA, on the other hand, will potentially contribute to other fields of study, including, importantly, linguistics. In this sense, MDA ‘use[s] texts or types of text to explore, illustrate, problematise, or apply general issues in multimodal studies, such as those arising from the development of theoretical frameworks specific to the study of multimodal phenomena, or methodological issues’ (O’Halloran and Smith, in press). This chapter deals with MDA precisely in this way − as a new field of study which requires specific theoretical and methodological frameworks and tools which in turn may be applied across other disciplines and domains.

Theoretical and Analytical Issues in MDA Theoretical and analytical issues in MDA include: (a) Modelling semiotic resources which are fundamentally different to language. (b) Modelling and analysing intersemiotic expansions of meaning as semiotic choices integrate in multimodal phenomena. (c) Modelling and analysing the resemioticization of multimodal phenomena as social practices unfold. These issues are considered in turn. (a) Modelling semiotic resources which are fundamentally different to language Following Halliday, language can be modelled as sets of interrelated systems in the form of system networks, which are metafunctionally organized according to taxonomies with hierarchical ranks (word, word groups, clauses, clause complexes and paragraphs and text (see Martin’s chapter in present volume). The grammatical systems link words to meaning on the semantic stratum (see Martin this volume). Systems which operate on the expression plane (i.e. graphology and typography for written language and phonology for spoken language) are also included in Halliday’s model. Most semiotic resources are fundamentally different to language, however, with those having evolved from language (e.g. mathematical symbolism, scientific notation and computer programming languages) having the closest relationship in terms of grammaticality. Images differ, for example, in that parts are perceived as organized patterns in relation to the whole, following Gestalt laws of organization. Furthermore, following Charles Sanders Pierce’s categorization of signs, language is a symbolic sign system which has

124

HylandK_Chap08_Finals.indd 124

12/1/2010 12:58:14 PM

Multimodal Discourse Analysis

no relationship to what is being represented, while images are iconic because they represent something though similarity. Therefore, analytic approaches and frameworks based on linguistic models have been questioned (Machin 2009). Nevertheless, models adapted from linguistics such as O’Toole (2010) have been widely and usefully applied to mathematical and scientific images, cities, buildings, museums and displayed art. In O’Toole’s model, the theoretical basis is Gestalt theory where images are composed of interrelated parts in the composition of the whole. O’Toole (2010) draws visual overlays of systemic choices on the image, suggesting a visually defined grammar as a possible way forward. Gestalt theory provides the basis for other approaches to visual analysis, including computational approaches to visual perception involving geometrical structures (e.g. points, lines, planes and shapes) and pattern recognition (e.g. Desolneux et al. 2008) and visual semantic algebras (e.g. Wang 2009). Perhaps one key to such descriptions is the provision of an abstract intermediate level, where low level features are related to semantics via systemic grammars. However, the problem is that hierarchically organized categorical systems such as those developed for language have limitations when it come to resources such as images, gestures, movement and sound which are topological in nature (Lemke 1998, 1999). Van Leeuwen (1999, 2009) proposes modelling systems within multimodal semiotic resources (e.g. colour, font style and font size for typography, and volume, voice quality and pitch) as sets of parameters with gradient values rather than categorical taxonomies ordered in terms of delicacy (i.e. subcategories with more refined options). In some cases, the existence of an intermediate grammatical level for resources such as music has been questioned (see van Leeuwen 1999). (b) Modelling and analysing intersemiotic expansions of meaning as semiotic choices integrate in multimodal phenomena The interaction of semiotic choices in multimodal phenomena gives rise to semantic expansions as the meaning potential of different resources are accessed and integrated; for example, in text–image relations (Bateman 2008; Liu and O’Halloran 2009; Martinec 2005; Unsworth and Cleirigh 2009) gesture and speech (Martinec 2004) and language, images and mathematical symbolism (Lemke 1998; O’Halloran 2008). This semantic expansion is also related to the materiality of the multimodal artefact, including the technology or other medium involved (e.g. book, interactive digital media) (Jewitt 2006; Levine and Scollon 2004; van Leeuwen 2005). Semantic integration in multimodal phenomena may be viewed metafunctionally whereby experiential, logical, interpersonal and textual meaning interact across elements at different ranks (e.g. word group and image). The

125

HylandK_Chap08_Finals.indd 125

12/1/2010 12:58:14 PM

Bloomsbury Companion to Discourse Analysis

resulting multiplication of meaning (Lemke 1998) leads to a complex multidimensional semantic space where there may be a compression of meaning (Baldry and Thibault 2006) and divergent (even conflicting) meanings (Liu and O’Halloran 2009). Indeed, there is no reason to assume a coherent semantic integration of semiotic choices in multimodal phenomena. The processes and mechanisms of semantic expansion arising from intersemiosis have yet to be fully theorized. It may be that intersemiotic systems beyond the sets of interrelated grammatical systems for each resource, operating as ‘meta-grammars’, are required. These intersemiotic systems would have the potential to link choices across the hierarchical taxonomies for each resource, so that a word group in language, for example, is resemioticized as a component of a complex visual narrative, or vice versa. One major problem for multimodal discourse analysts is the complexity of both the intersemiotic processes and the resulting semantic space, particularly in dynamic texts (e.g. videos) and hypertexts with hyperlinks (e.g. internet). (c) Modelling and analysing the resemioticization of multimodal phenomena as social practices unfold MDA is also concerned with the resemioticization of multimodal phenomena across place and time: ‘[r]esemioticisation is about how meaning making shifts from context to context, from practice to practice, or from stage of a practice to the next’ (Iedema 2003: 41). Iedema (2003: 50) is concerned with resemioticization as a dynamic process which underscores ‘the material and historicised dimensions of representation’. Resemioticization takes place within the unfolding multimodal discourse itself (as the discourse shifts between different resources) and across different contexts as social practices unfold (e.g. how a policy document is enacted). From a grammatical perspective, resemioticization necessarily involves a reconstrual of meaning as semiotic choices change over place and time. In many cases, resemioticization involves introducing new semiotic resources, and may result in metaphorical expansions of meaning as functional elements in one semiotic resource are realized using another semiotic resource: for example, the shift from language, to image and mathematical symbolism in unfolding mathematics discourse. This process takes place as linguistic configurations involving participants, processes and circumstances, for example, are visualized as entities. Resemioticization necessarily results in a semantic shift, as choices from different semiotic resources are not commensurate (Lemke 1998). Processes specific to MDA, such as intersemiosis and resemioticization of multimodal phenomena, add to the complexity of the semantic space which must be modelled and analysed. Indeed, managing this complexity lies at the heart of MDA.

126

HylandK_Chap08_Finals.indd 126

12/1/2010 12:58:14 PM

Multimodal Discourse Analysis

Sample MDA Text Analysis Concepts specific to MDA, namely semiotic resource, intersemiosis and resemioticization, are illustrated through the analysis of an extract from a television multiparty debate, Episode Two of the Australian Broadcasting Commission’s (ABC) television show ‘Q&A: Adventures in Democracy’ broadcast on Thursday 29 May 2008. The moderator is senior journalist Tony Jones and the panel consists of Tanya Plibersek (Minister for Housing and the Status of Women in Kevin Rudd’s Federal Labor Government), Tony Abbott (then Opposition Liberal Party front-bencher, now Leader of the Opposition in the Australian House of Representatives) and Bob Brown (Leader of the Australian Green Party). Other participants in the panel discussion, although not considered here, are Warren Mundine (Indigenous Leader and former president of the Australian Labor Party) and Louise Adler (CEO and Publisher-in-Chief of Melbourne University Publishing).1 The extract is concerned with interactions between Tony Jones, Tanya Plibersek and Tony Abbott about leaked cabinet documents regarding a Government Cabinet decision in favour of a Fuel-Watch scheme to combat rising petrol prices, and reservations about this scheme as revealed through the leaked documents. (Note: * indicates overlap). Tanya Plibersek

. . . The reason that cabinet documents are confidential is that so senior public servants feel comfortable giving frank advice to the government of the day.

Tony Jones

Alright. Tony Abbott, you’ve been in the trenches. That’s fair enough isn’t it.

Tony Abbott:

Ah, yes it is, but the interesting thing is that the new government is already leaking Tony. I mean normally it takes many years *before a – before – before a government . . . well I -

Tony Jones:

* yes a little – a little bit like the coalition. Leaking going on all round.

Tony Abbott:

Tired old governments leak. New, smart, clever, intelligent governments aren’t supposed to leak, and the fact that this government is leaking so badly so early is a pretty worrying sign.

The multimodal analysis includes the interactions between the spoken language, kinetic features (including gaze, body posture and gesture) and cinematography effects (including camera angle and frame size) (see also Baldry and Thibault 2006; Iedema 2001a; Tan 2005, 2009). The multimodal analysis 127

HylandK_Chap08_Finals.indd 127

12/1/2010 12:58:14 PM

Bloomsbury Companion to Discourse Analysis

presented here is for illustrative purposes only. A more comprehensive linguistic analysis could have been presented, in addition to the inclusion of other semiotic resources (e.g. studio lighting, clothing, proxemics, seating arrangement and so forth). Furthermore, semiotic choices are presented in a static table (see Table 8.2), rather than a dynamic format which would have permitted the unfolding of choices and patterns to be represented. Halliday’s (2004; Halliday and Greaves 2008) systemic functional model for language (including intonation) and Tan’s (2005, 2009) systemic model for gaze and kinetic action (Figure 8.1) and camera angle, camera movement, and visual frame (Table 8.1) are drawn upon for the analysis, as is van Leeuwen’s work on the semiotics of speech rhythm (e.g. 1999). Comprehensive descriptions of these models are found elsewhere, and thus are not repeated here. The multimodal analysis of the extract with key salient frames are presented in Table 8.2. The following analysis reveals how the multimodal choices Tony Abbott makes, particularly with respect to linguistic choices, intonation, gesture and body posture, work closely together to reorientate the discussion about the leaked documents from being a legal issue to a political issue in order to criticize and undermine Kevin Rudd’s (the former Australian Prime Minister) Labor government.

Table 8.1 Camera Angle, Camera Movement and Visual Frame (Tan 2009: 179) Angle/Power, Perspective HP VP POV

Horizontal Angle: frontal angle signals involvement, oblique angle signals detachment Vertical Angle denotes power relations: high/median/low Point-of-View (subjective image) Camera Movement

CM stat mobile dolly pan tilt zoom-in/out ³´µ¶·¸ ¹º 34

Camera Movement Stationary Camera Mobile Framing Camera travels in any direction along the ground: forward, backward, circularly, diagonally, or from side to side Camera scans space horizontally from left to right or right to left Camera scans spaces vertically up or down Camera does not alter position; space is either magnified or de-magnified Directionality of camera movement is indicated by short directional arrows Size of Visual Frame

close-up extreme close-up extreme long shot long shot medium long shot medium shot medium close-up

Shows just the head, hands, feet, or a small object Singles out a portion of the face (eyes or lips) Human Figure is barely visible; landscapes, bird’s-eye views Full view of human figure(s) with background Human Figure is framed from about the knees up Frames the human body from the waist up Frames the body from the chest up

128

HylandK_Chap08_Finals.indd 128

12/1/2010 12:58:14 PM

HylandK_Chap08_Finals.indd 129

eye contact directed at other participant(s) Actor and Goal connected by Vector = connected transaction (material process)

body parts clothing, props

inside personal space engaged

Diegetic world/ on-screen space

directed at self

disengaged Directionality of Gaze and Kinetic Action Vectors

directed at object(s) Reactor and Phenomenon connected by vector (= transitive reaction)

outside personal space

body parts

clothing, props no vector, or movement only (intransitive, material action process) self-involvement (Mental Process)

Interpersonally engaged; gaze directed at viewer = direct address/visual demand

Off-screen space

engaged; participant is Reactor, Phenomenon shown in next Shot = disconnected transactive reaction (realized through POV-shots, shot/shot-reverse shots, eyeline matches, match-on-action shots)

12/1/2010 12:58:14 PM

disengaged; only one participant, no Phenomenon = non-transactive reaction (Mental Process)

Figure 8.1 Systemic networks for Gaze and Kinetic Action Vectors (Tan 2005: 45)

no clear Vector = Circumstance of Means no Vector = Circumstance of Accompaniment

HylandK_Chap08_Finals.indd 130

Table 8.2 Multimodal analysis of ‘leaked cabinet documents’ (Q&A Session, ABC Thursday 29 May 2008) Stage Phase Sub-Phase

Leaking Documents as legal Issue SHOT 1 SHOT 2

‘Petrol Prices’ Leaked Cabinet Documents Leaking Documents as Political Issue SHOT 3

SHOT 4

SHOT 5

Salient Visual Frame

Frame 1

Frame 2

Frame 2

Frame 4

Frame 5

Frame 6

*yes a little – a little bit like the coalition.

Leaking going on all round

Frame 7

Frame 8

Frame 9

Frame 10

leak. and the fact that this government is leaking so badly

so early is a pretty worrying sign.

SEMIOTIC RESOURCE: Speech: Speaker 1Tony Jones (interviewer):

Alright Tony Abbott

you’ve been in the trenches. That’s fair enough isn’t it?

Speaker 2Tony Abbott: Ah, yes it is

but the interesting thing is that the new government is already

leaking Tony. I mean normally it takes many years

*before a – before – before a government

... well I – Tired old governments

leak. New, smart, clever.

intelligent governments aren’t supposed to

off-screen; engaged; directed at interviewer

off-screen; engaged; directed at Tony Abbott

off-screen; disengaged; directed at self

off-screen; engaged; directed at studio audience//interviewer/Tanya Plibersek

off-screen; engaged; directed at studio audience//interviewer/Tanya Plibersek

off-screen; engaged; directed at camera/viewer

off-screen; engaged; directed at studio audience//interviewer/Tanya Plibersek

off-screen; engaged; directed at studio audience//interviewer/Tanya Plibersek

angled

angled leans forward toward Tony Abbott

angled leans back

angled

angled

straight

angled

angled

raises hand; palm facing outward

raises hand; palm facing outward

hand raised; palm facing outward

both hands raised; palms facing outward/each other

Kinetic Features: Gaze:

Body Posture:

off-screen; engaged; directed at Tony Abbott

angled

Gesture:

off-screen; engaged; directed at studio audience//interviewer/Tanya Plibersek angled

both hands raised; palms facing outward/each other at reduced distance; downward movement

both hands raised; both hands raised; palms facing palms facing outward/each outward/each other; gap other; gap narrowing narrowing

Cinematography:

12/1/2010 12:58:15 PM

Camera Angle (horizontal perspective)

oblique/ detached

oblique/ detached

oblique/ detached

oblique/ detached

oblique/ detached

frontal/ involved

oblique/ detached

oblique/ detached

Size of Frame

medium close-up

medium close-up

medium close-up

medium close-up

medium close-up

medium close-up

medium close-up

medium close-up

oblique/ detached medium close-up

oblique/ detached medium close-up

Multimodal Discourse Analysis

Figure 8.2 Tony Abbot’s use of Tone 4 (Halliday and Greaves 2008) in ‘It IS . . . ’ (Image produced using Praat software)

Tony Jones puts forward to Tony Abbott a proposition with the tag ‘isn’t it’ (which explicitly signals that a particular kind of response is required) with respect to Tanya Plibersek’s defence of her government’s handling of the leaked documents: ‘That’s fair enough isn’t it?’ The (exaggerated) tone 4 (fall-rise) of Tony Abbott’s reply ‘Ah, yes it is . . . ’ (displayed in Figure 8.2) adds reservation to this proposition, and is an interpersonally focused reply, both in the sense of having the information focus on the Finite ‘is’ – the negotiatory element of the clause – but also in that there is no addition of experiential meaning (in terms of content), until Tony Abbott continues with ‘but the interesting thing is that the new government is already leaking Tony’. Tony Abbott thus concedes (via polarity) the proposition as put, but enacts reservation (via intonation) with respect to another field of discourse, that of politics: that the new government is already leaking. Thus for him the legal issue is not what is at stake here, rather there is a shift to the leaking of the documents as a political issue, resulting in a new sub-phase in the Leaked Cabinet Documents phase (see Table 8.2 and Figure 8.3(a)). He moves the battle to a new ground, and then proceeds to elaborate on his point.

131

HylandK_Chap08_Finals.indd 131

12/1/2010 12:58:16 PM

Bloomsbury Companion to Discourse Analysis

Figure 8.3(a) The change of field from legal issue to political issue

Figure 8.3(b) Gaze and gesture

132

HylandK_Chap08_Finals.indd 132

12/1/2010 12:58:17 PM

Multimodal Discourse Analysis

Figure 8.3(c) Body posture Figure 8.3 Tony Abbott’s ‘leaking documents’ as political issue

This shifting of the field of discourse is a characteristic of political discourse (well known as ‘politicians not answering the question’) but in this case, it is possible to see how Tony Abbott effectively employs a range of multimodal resources which function intersemiotically to change the field of discourse, displayed in Table 8.2 and Figure 8.3(b)-(c). These resources include clause grammar (adversive conjunction ‘but’); information unit grammar (use of the ‘reserved’ key, realized through falling-rising tone 4); gesture (holding up his hand in a ‘wait on’ movement, which then becomes the preparation for a series of gesture strokes to emphasize the points made, see Figure 8.3(b)); body posture (first, sitting back and then leaning forward as he makes his point about the new government leaking); and interpersonal deixis (vocative ‘Tony’ enacting solidarity). Following this, Tony Abbott continues speaking as he sits back and then engages successively with the studio audience, Tony Jones and Tanya Plibersek through gaze and angled body posture, while expanding his hand gesture somewhat (see Figure 8.3(b)-(c)). He also briefly but directly engages with the viewer with a straight body posture with both hands raised and palms facing outwards to further engage the viewer, before turning his attention back to the panelists Tanya Plibersek and Tony Jones and the studio audience. Tanya Plibersek’s ‘nonplussed’ response in the form of gaze and facial expression (Frame 9 in Table 8.2, also see second last frame in Figure 8.3)) is a study in itself: she makes no other significant semiotic sign, but is clearly quite familiar with her political opponent’s stratagems. Note that the camera is deployed as a semiotic resource here, in the choice to frame her at this point, setting up a dialogic context between Tony Abbott and herself, despite the fact that it was Tony Jones who asked the question. Tony Abbott uses gesture and speech rhythm to emphasize lexical items, raising the textual status both of the individual words themselves and the overall point and thereby creating a form of a graduation in emphasis (Martin 133

HylandK_Chap08_Finals.indd 133

12/1/2010 12:58:18 PM

Bloomsbury Companion to Discourse Analysis

and White 2005). The use of gesture and accent together provide a more delicate range of textual gradience, organizing the flow of information into varying degrees of prominence – a semiotic expansion arising from the combined visual and aural gradience of the bandwidths of gestural stroke and accent. At this critical point Abbott establishes a crucial intertextual reference (Lemke 1995) to the whole discourse of the previous Federal election in Australia, when his Liberal government of 11 years was soundly defeated by an opposition which projected itself as being fresh and ‘clever’ by contrast with the ‘tired, old’ incumbent government. He does this primarily through rhythm: up to the point where he says ‘tired, old governments leak’ he sets up a distinct temporal patterning of accents, which is then disturbed at the point between ‘clever’ and ‘intelligent’ in ‘New, smart, clever, intelligent governments aren’t supposed to leak.’ Abbott thus plays ironically here on this recent electioneering discourse – and his direct gaze (see Frame 8 in Table 8.2) also takes on a semiotic rendering of the ironic satirical tone, as a visual signal of ‘playing it straight’. There are many other opportunities to demonstrate how multimodal resources function intersemiotically to achieve the agenda of the involved parties, including the producers who use camera shots to create a dialogue between the participants. For example, while Tony Jones engages Tanya Plibersek in a critical dialogue about a Government environmental policy initiative, the camera view changes to include Bob Brown, Leader of the Australian Green Party, who is seen to raise his eyebrows, nod his head, lick his lips and shake his head from side to side, which gestures, afforded by choice of camera shot, entirely recontextualizes the dialogue of which Brown at this point is not (verbally) a part (see Figure 8.4). The entire Q&A session itself is resemioticized on the Q&A website (Figure 8.5) where the notion of political debate as sport is evoked in the

Figure 8.4 Camera: Visual Frame

134

HylandK_Chap08_Finals.indd 134

12/1/2010 12:58:18 PM

Multimodal Discourse Analysis

Figure 8.5 Q&A website: adventures in democracy – ‘Tony, Tanya and Bob’ (retrieved from http://www.abc.net.au/tv/qanda/txt/s2255680.htm)

opening paragraph (‘Tony, Tanya and Bob. Thursday, 29 May. Tony Abbott and Tanya Plibersek are back in the boxing ring for Q&A’s second episode. Joining them are Bob Brown, Warren Mundine and Louise Adler for their first grilling by the Q&A punters’). But the ‘spectators’ – the audience – are encouraged to participate, through interactive blog forums arrayed under each of the show’s questions where website members may post comments (‘Have your say’), another resemioticization of the issues debated during the show (from expert to public opinion), as well as post questions for the show itself (including ‘live’ questions during the show). A mathematical chart post-show also gives some (limited) analytical information about the time devoted to the topics under discussion, and further down the website the panelists are introduced via photos and short write-ups. The above discussion shows clearly that context is an essential part of any analysis, not just the immediate context of situation (the Q&A event and subsequent resemioticizations of that event), but the context of culture in general, including in this case the intertextual references which are made to the recent elections in Australia and its discourse, and to Australian democratic culture in general. MDA reveals how instances of multimodal semiotic choices function

135

HylandK_Chap08_Finals.indd 135

12/1/2010 12:58:19 PM

Bloomsbury Companion to Discourse Analysis

intersemiotically in ways which ultimately create and answer to larger patterns of social context and culture.

New Directions in MDA The major challenge to MDA is managing the detail and complexity involved in annotating, analysing, searching and retrieving multimodal semantics patterns within and across complex multimodal phenomena. The analyst must take into account intersemiotic and resemioticization processes across disparate timescales and spatial locations. In addition, different media may require different theoretical approaches, for example, video and film analysis may draw upon insights from film studies (Bateman 2007). MDA of websites and hypermedia give rise to added difficulties as semiotic choices combine with hypermedia analysis of links and other navigational resources, resulting in hypermodal analysis (Lemke 2002). One method for managing the complexity involves the development of interactive digital media platforms specifically designed for MDA. Furthermore, the development of software as a metasemiotic tool for multimodal analysis becomes itself a site for theorizing about and developing MDA itself. Multimodal annotation tools currently exist (Rohlfing et al. 2006), while further work is underway to develop interactive software for MDA which goes beyond annotation to include visualization and mathematical techniques of analysis (O’Halloran et al. 2010). The path forward must necessarily involve interdisciplinary collaboration if the larger goals of understanding patterns and trends in technologies, text, context and culture are to be achieved.2

Notes 1. My sincere thanks to Bradley Smith and Sabine Tan from the Multimodal Analysis Lab, Interactive and Digital Media Institute (IDMI) at the National University of Singapore for their significant contributions to the Q&A analysis. Also, thanks to Bradley Smith for providing the Q&A extract and Figure 8.2. 2. Research for this article was undertaken in the Multimodal Analysis Lab IDMI at the National University of Singapore, supported by Media Development Authority (MDA) in Singapore under the National Research Foundation’s (NRF) Interactive Digital Media R&D Program (NRF2007IDM-IDM002–066).

Key Readings Halliday, M. A. K. (1978), Language as Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold.

136

HylandK_Chap08_Finals.indd 136

12/1/2010 12:58:20 PM

Multimodal Discourse Analysis Jewitt, C. (ed.) (2009), The Routledge Handbook of Multimodal Analysis. London: Routledge. Kress, G. and van Leeuwen, T. (2006), Reading Images: The Grammar of Visual Design (2nd edn). London: Routledge. Norris, S. (2004), Analyzing Multimodal Interaction: A Methodological Framework. London: Routledge. O’Toole, M. (2010), The Language of Displayed Art (2nd edn). London and New York: Routledge.

137

HylandK_Chap08_Finals.indd 137

12/1/2010 12:58:20 PM

9

Corpus Approaches to the Study of Discourse Bethany Gray and Douglas Biber

Chapter Overview Introduction Researching Language in Use through Corpus Analysis A Sample Study: Language in Use Investigating Language Structure beyond the Sentence and Future Directions for Corpus-based Discourse Analysis Note Key Readings

138 139 143 151 152 152

Introduction In this chapter, we describe how corpus linguistics can be used to analyse discourse from two perspectives: language in use and language structure beyond the sentence. While ‘discourse’ has been conceptualized in many different ways (see Jaworski and Coupland 1999; Schiffrin 1994), Schiffrin et al. (2001: 1) state that the approaches to discourse can be grouped into three major categories (see Biber et al. 2007: Ch. 1 for further discussion): (1) discourse as language in use, which investigates variation in the use of linguistic forms and traditional linguistic constructs; (2) discourse as language structure above the sentence level, which focuses on the broader text structure, that is, on the systematic ways that texts are constructed; and (3) discourse as social practices and ideologies associated with language and/or communication, focusing on the general characteristics and participants of a particular discourse community. 138

HylandK_Chap09_Finals.indd 138

12/1/2010 8:04:38 AM

Corpus Approaches to the Study of Discourse

In the first two perspectives, there is an explicit focus on linguistic form in a text or collection of texts; thus, as a method for describing linguistic form, corpus linguistics is particularly applicable to studying discourse from these two perspectives. Corpus linguistics is a methodology for linguistic analysis that focuses on describing linguistic variation in large collections of authentic texts (the corpus), using automatic and interactive computer programs to aid in analysis. Thus, the goal of corpus linguistics is to identify patterns of variation that are generalizable across many texts in a specific discourse context or language variety. This variation can occur at many levels, and the wide variety of linguistic features being examined through corpus linguistics methodologies today illustrates this. For example, corpus linguistics research can examine the choice between semantically similar words (e.g. little, small), or the choice between nearly synonymous grammatical structures (e.g. active versus passive voice, dative movement, particle movement in phrasal verbs). Going beyond descriptions of a particular linguistic feature, corpus linguistics can also document the relative distributions of many lexical and grammatical features in any particular variety of language use, leading to comparison between varieties. Because most corpus-based studies focus on linguistic forms in specific linguistic and/or situational contexts, corpus linguistics is primarily seen as belonging to the language in use approach, and this area will be the focus of the present chapter. However, as we will discuss in the final section, corpus linguistics can also be used to study text structure. The goal of the present chapter is to introduce corpus linguistics as a method of discourse analysis, demonstrating the distinctive insights corpus analysis can provide into how a particular language variety is constructed and realized linguistically. Thus, in the following section, we describe the key characteristics of corpus linguistics methodologies in relation to its use for discourse analysis and summarize some of the major approaches taken to discourse analysis within corpus linguistics. In the third section, we provide a case study investigating structural elaboration in academic writing and conversation, illustrating a corpus-based approach to studying discourse as language in use. In the final section, we address further directions for corpus-based discourse analysis, extending our discussion to corpus-based studies of discourse that consider text structure (discourse as language beyond the sentence).

Researching Language in Use through Corpus Analysis Over the past few decades, the development of corpus linguistics as a methodological approach to descriptive linguistic analysis has provided a means of making generalizable discoveries about language use in a variety of situations. 139

HylandK_Chap09_Finals.indd 139

12/1/2010 8:04:39 AM

Bloomsbury Companion to Discourse Analysis

These discoveries are based on the empirical evidence provided by large collections of naturally occurring texts. As a methodological approach, corpus linguistics has the following characteristics (Biber et al. 1998: 4): 1. The approach is empirical. Findings are based on actual, observed patterns of use in naturally occurring language. 2. The foundation for the analysis is the corpus, a principled collection of natural texts that are representative of a target domain. 3. Computers are used to analyse the corpus through both automatic and interactive tools. 4. The analysis is both quantitative and qualitative in nature. The major goal of most corpus-based studies of discourse is to document variation in the use of lexical and grammatical features in particular registers, describing the functions and frequency distributions of those linguistic features. The four characteristics introduced above distinguish corpus linguistics methodologies from other discourse analysis approaches, and provide the foundation for the unique contributions that corpus studies can offer.

The Strengths of Corpus-based Research First, the analysis is based on a corpus, a large collection of electronic texts that is sampled to represent a target register, or language variety. While many terms have been used to distinguish varieties of language, including register, genre, style and text type,1 we use the term register to refer to varieties of language use that are distinguishable based on situational characteristics such as purpose, mode, setting, author/speaker, reader and so on (see Biber and Conrad 2009). For example, we can distinguish conversation as a register by its real time spoken mode, shared time and place between participants, and highly interactive nature, among others. Academic writing, on the other hand, is defined by its written mode, the use of editing practices, the lack of shared time and place among participants, and its highly informational purpose. Because the corpus is specifically designed to represent the target domain, findings can be meaningfully generalized to the register as a whole. Rather than an analysis of a single or small number of texts, corpus studies can examine a large collection of texts efficiently. Second, corpus-based language studies rely on a variety of computer tools to assist in the analysis of the corpus. Because a computer should come to the same conclusion about a particular linguistic feature each time it is encountered, the reliability of computer-aided analysis is typically quite high, allowing the researcher to focus attention on the interpretation of the linguistic data. 140

HylandK_Chap09_Finals.indd 140

12/1/2010 8:04:39 AM

Corpus Approaches to the Study of Discourse

In addition and perhaps even more importantly, the efficiency afforded by computers means that a much greater amount of language data can be analysed. As a result of the extensive amount of language data available, calculating quantitative patterns of use (in addition to qualitative analyses) is meaningful and in fact, fairly easy and reliable to carry out. The use of quantitative data in corpus-based discourse analysis is important for two main reasons. First, quantitative methods allow for the empirical establishment of linguistic patterns across the texts in the corpus. That is, the researcher can identify patterns of variation that are consistent across all the texts in the corpus, reducing the risk that an analysis will be an inaccurate description of a register as a whole because the texts analysed were not actually representative of that register. Second, because of the ability to describe a register as a whole (in quantitative terms and based on a large number of representative texts), it becomes quite feasible to make comparisons across registers. Of course, a register cannot be fully described in quantitative terms alone. Key to the application of corpus linguistics is the belief that language variation is both systematic and functional. That is, corpus linguistics research recognizes the inherent linguistic variation that occurs due to differences in situations of language use. While many linguistic and situational factors may interact to create variation, that variation is systematic nonetheless and can be explained by considering the functional characteristics of the situation of use and the linguistic features themselves. Thus, a second key goal of corpus linguistics research is to explain the quantitative patterns of variation in relation to the functions of the linguistic features and how those functions match the needs of a communicative situation (as illustrated in the sample study below). In summary, the core contribution of corpus linguistic approaches to discourse analysis is the ability to make reliable, generalizable discoveries about the patterns of linguistic variation across registers. Because the corpus is central to the generalizability of corpus research findings, the design of the corpus is a crucial consideration for the researcher. More specifically, two characteristics of the corpus are of particular concern for corpus linguists: size and representativeness (see Biber 1993; Biber et al. 1998; McEnery et al. 2006: 13–21, 125–30). In terms of size, the corpus must be large enough to provide a sufficient number of occurrences of the linguistic feature(s) of interest, and the appropriate size varies depending on the nature of the linguistic features, with studies of lexical variation generally requiring larger corpora than studies of grammatical variation. In terms of representativeness, the key concern is that the texts included in the corpus are sampled in a manner that ensures that they accurately represent texts in the register/ domain of use. 141

HylandK_Chap09_Finals.indd 141

12/1/2010 8:04:39 AM

Bloomsbury Companion to Discourse Analysis

The Nature of Corpus-based Studies of Language in Use Within the general framework of language in use, corpus-based discourse studies have investigated a wide variety of linguistic features. One general line of inquiry seeks to describe the use of specific linguistic features within and across registers. For example, corpus-based studies have investigated aspects of lexis, such as studies of collocation (the co-occurrence of particular words, for example, Gledhill 2000; Ward 2007), studies identifying the statistically distinctive lexical items in a register (keyword analysis, see for example, Freddi 2005; Scott and Tribble 2006), and studies of lexical bundles, which are frequently occurring multi-word sequences (see Biber et al. 1999, chapter 13; Cortes 2004; Kim 2009; and Nesi and Basturkmen 2006). In addition, researchers are increasingly relying on corpus linguistics methods to uncover distributional patterns of grammatical features in discourse, as shown by the many book-length treatments reporting on corpus studies of grammar and discourse: for example, Collins (1991) on clefts, Hunston and Francis (2000) on pattern grammar, and Römer (2005) on progressives. While most of these studies focus on a single linguistic feature or construct, other research focuses on analysing collections of feature which contribute to a single function (e.g. Hyland 1998 and Biber 2006b on stance; Afros and Schryer 2009 on self-promotion). However, all of these types of research focus on describing the functions and distributions of individual linguistic features, or small sets of related features, within and across registers. A second line of inquiry seeks to describe not a particular language feature, but rather to describe registers in terms of the co-occurrence patterns of many linguistic features. This line of research follows a specific methodology, called multidimensional analysis (see Biber 1988; Conrad and Biber 2001). In multidimensional analysis, the statistical technique of factor analysis is used to analyse the frequency distributions of many linguistic features in the corpora to locate those that frequently co-occur. After the statistical analyses, qualitative analyses of the functions of the co-occurring features is undertaken to interpret the findings, leading to generalizations about ‘dimensions’ of language use. Multidimensional analysis has been used to investigate many different registers, from spoken and written academic texts (Biber et al. 2002) to the language of international call centres (Friginal 2009) to television dialogue (Quaglio 2009). Corpus-based approaches to linguistic analysis are ideally suited to comparisons of the use of linguistic constructs across registers, in part because of the precise quantitative data yielded from corpus studies and because of the ability to collect large, representative corpora of a variety of registers. Corpus-based research as a whole has shown that differential patterns

142

HylandK_Chap09_Finals.indd 142

12/1/2010 8:04:39 AM

Corpus Approaches to the Study of Discourse

of variation across registers exist for a single linguistic feature as well as sets of co-occurring features. That is, ‘characteristics of the textual environment interact with register differences, so that strong patterns in one register often represent weak patterns in other registers’ (Biber 2010: 162). Thus, corpus-based research has shown that linguistic description which does not account for variation across registers is generally not a complete or necessarily accurate description of the language as a whole. For this reason, many corpus-based studies consider register differences in their description of the distribution and functional uses of linguistic constructs. For example, the Longman Grammar of Spoken and Written English (Biber et al. 1999) is a comprehensive corpus-based reference grammar that documents how any grammatical feature of English is distributed and used across a variety of spoken and written registers. Although a few corpus studies focus on describing the variation of a particular feature within a single register, most corpus studies compare the use of a feature across multiple varieties. Identifying the distinctive use of features in a register necessarily requires that the researcher compare the use of that feature in one register to its use in another register. These comparisons are realized in a variety of ways. For example, many times very diverse registers are compared, as is the case in comparisons of speech and writing (e.g. Biber and Gray 2010; Biber et al. 2002). Other times, a particular ‘macro’ register like academic writing is being considered, with comparisons being made across academic disciplines (e.g. Hyland 2001a, b; Hyland and Tse 2004), across different types of texts within academic writing (e.g. Conrad’s 1996 comparison of research articles, textbooks and student writing), or across specific sections of a target register (e.g. Martínez 2005; Chen and Ge 2007). As we will explain below, even when language beyond the sentence is investigated through corpus analysis, the approach is comparative; the use of linguistic features is compared across moves types.

A Sample Study: Language in Use The following study (adapted from Biber and Gray 2010) compares the use of linguistic features associated with structural complexity and elaboration in conversation and academic writing. This study illustrates a language in use research approach, and highlights the importance of making register comparisons in corpus-based research. In addition, this study demonstrates how register-based comparisons can reveal patterns of variation counter to language users’ and researchers’ intuitions and assumptions about how language is used in a particular register.

143

HylandK_Chap09_Finals.indd 143

12/1/2010 8:04:40 AM

Bloomsbury Companion to Discourse Analysis

Complexity and Elaboration in Academic Prose and Conversation Researchers have long claimed that in comparison to conversation, academic writing is structurally elaborated, relying more heavily on subordinate clauses (e.g. see Brown and Yule 1983; Chafe 1982). Hughes (1996: 58–64) notes that spoken grammar employs ‘simple and short clauses, with little elaborate embedding’, in direct contrast to writing, which uses ‘longer and more complex clauses with embedded phrases and clauses’. A few earlier studies have shown that academic writing relies heavily on nominal structures (e.g. Halliday and Martin 1993/1996; Wells 1960), and that clausal subordination is actually more common in speech while phrasal embedding is more common in writing (e.g. Biber 1988, 1992, 2006a: chapter 7). However, despite these studies, the belief that academic writing is structurally elaborated and employs extensive use of clausal embedding (in contrast with speech) persists today for both language teachers and researchers. Hyland (2002: 50) also documents this popular perception of academic writing as ‘structurally elaborate, complex’ and with ‘more subordination’. In sum, academic writing is generally seen as more complex and elaborated than speech, particularly in the use of traditional complexity measures such as embedded dependent clauses. Corpus-based linguistics is an ideal method for investigating these widely held notions about speech and academic writing. By comparing the distributions of linguistic features related to structural elaboration in corpora representing the two target domains of speech and academic writing, we can see how the actual relative distributions of these features differ across the two registers, and how those distributions are contrary to widely held beliefs about conversation and academic prose.

Method and Corpus The first step in the analysis is to identify linguistic features which are related to structural elaboration. Table 9.1 lists five of these features identified from previous research. All five structures are dependent clause types. Complement clauses (finite and non-finite) are usually functioning as the direct object of a verb, and include that-clauses, wh-clauses and to-clauses. Adverbial clauses modify the main verb of the sentence, and relative clauses typically occur after the head noun they modify. These features are considered ‘elaborated’ features. Complement clauses typically function as the direct object of a verb, but they are elaborated in the sense that they use a clause to fill a sentence slot normally filled by a noun phrase. Adverbial clauses and relative clauses, on the other hand, are

144

HylandK_Chap09_Finals.indd 144

12/1/2010 8:04:40 AM

Corpus Approaches to the Study of Discourse

Table 9.1 Grammatical features associated with structural elaboration Grammatical Feature

Examples

Finite complement clauses

I don’t know how they do it. I thought that was finished. I’d like to get one of these notebooks. Do you want to elaborate on that more? She won’t narc on me, because she prides herself on being a gangster. So she can blame someone else if it doesn’t work. the quantity of waste that falls into this category . . . There are three sets of conditions under which the crop is raised. The results shown in Tables IV and V add to the picture . . . The factors contributing to the natural destruction of microbes . . .

Non-finite complement clauses Finite adverbial clauses

Finite relative clauses

Non-finite relative clauses

Table 9.2 Grammatical features associated with structural compression Grammatical Feature

Examples

Attributive adjective (adjective as noun pre-modifier) Noun as noun pre-modifier Appositive noun phrase as noun postmodifier Prepositional phrase as noun postmodifier

a large number, unusual circumstances

Prepositional phrase as adverbial

Alright, we’ll talk to you in the morning. Is he going to the store?

surface tension, liquid manure In four cohorts (Athens, Keio, Mayo, and Florence), investigators stated that . . . Class mean scores were computed by averaging the scores for male and female target students in the class. Experiments have been conducted to determine the effect of salt on the growth and development of paddy.

elaborated in the sense that they are optional additions to the main clause, added on in order to provide elaborating information to the main verb or head noun. As a point of comparison, we also investigate a set of features which are not ‘elaborated’ structures, but rather serve to ‘compress’ information into dense units. These features are summarized in Table 9.2, with examples. Although all of these features also add additional information to either a head noun or the main clause, they are condensed or compressed structures used to pack information into a small number of words. This ‘compression’

145

HylandK_Chap09_Finals.indd 145

12/1/2010 8:04:40 AM

Bloomsbury Companion to Discourse Analysis

is evidenced by the fact that many of these structures can be rephrased with fuller structures, usually through the use of clausal modifiers; for example: unusual circumstances → liquid manure → effect of salt → students in the class →

circumstances which are unusual manure that is in a liquid state the effect which is caused by salt students who are taking the class

The next step in the analysis is to analyse the distributions of these features in corpora representing the target domains. For this analysis, we compared two corpora: (1) an academic writing corpus composed of 429 research articles (c. 3 million words) sampled from science/medicine, education, social science (psychology) and humanities (history) and (2) the conversation subcorpus (c. 4.2 million words of American English) from the Longman Spoken and Written Corpus (see Biber et al. 1999: 24–35). These corpora were annotated for grammatical categories (‘tagged’) using software developed for the Longman Grammar of Spoken and Written English (Biber et al. 1999). This process adds a ‘tag’ to each word in the corpus stating its grammatical part of speech. Then, more specialized computer programs were developed to analyse the features associated with structural elaboration (see Table 9.1) and structures related to structural compression (see Table 9.2). One of the defining characteristics of the corpus linguistics methodology is the reliance on both automatic and interactive computer techniques. While many of the features identified in Tables 9.1 and 9.2 could be identified automatically based on the grammatical annotation, determining the function of features like prepositional phrases (functioning as a noun postmodifier versus an adverbial) was an interactive process in which a computer program located potential occurrences of these features. These occurrences were then coded by hand (based on a subsample of every fourth token). Each of the features identified in Tables 9.1 and 9.2 were counted, and then normalized to counts per 1,000 words (see Biber et al. 1998: 263–4). Normalizing the counts to a standard reference frame allows for the direct comparison of the frequencies of these features across the two corpora, which contain differing total number of words.

Results of Corpus Analysis Commonly held perceptions that academic writing is more structurally elaborated than conversation would predict that academic writing would use the elaborated grammatical features to a greater extent than conversation. In fact, 146

HylandK_Chap09_Finals.indd 146

12/1/2010 8:04:40 AM

Corpus Approaches to the Study of Discourse 16

Conversation

Academic writing

14

Rate per 1,000 words

12

10

8

6

4

2

te r cl elat au iv se e s

ni

te r cl elat au iv se e s N

on

-fi

ni Fi

ad v cl erb au ia se l s

te ni Fi

N

on

-fi

Fi

ni

ni

te

te

co

co

m pl cl em au en se t s

m pl cl em au en se t s

0

Figure 9.1 Common dependent clause types (from Biber and Gray 2010)

the results show that the opposite is true. Figure 9.1 shows the distribution of common dependent clause types. Finite complement clauses (e.g. that-clauses and wh-clauses), non-finite complement clauses (to-clauses and ing-clauses) and finite adverbial clauses (e.g. because-clauses and if-clauses) are all much more prevalent in conversation than in academic writing. The two other types of dependent clauses – finite and non-finite relative clauses – are more prevalent in academic writing than conversation; these are both nominal features that modify a head noun. Overall, dependent clause types are nearly twice as frequent in conversation as in academic writing. In contrast, Figure 9.2 shows that much of the embedding and elaboration in academic writing comes from phrasal components modifying a head noun, with all four of the dependent phrasal structures occurring at a higher rate per 1,000 words in academic writing than in conversation. Attributive adjectives (e.g. differential reinforcement, theoretical orientation) and nouns as noun modifiers (e.g. trait information, system perspective) are quite common in academic prose in comparison to conversation. The distribution of prepositional phrases as noun postmodifiers (e.g. a strategic approach to mutual understanding; 147

HylandK_Chap09_Finals.indd 147

12/1/2010 8:04:41 AM

Bloomsbury Companion to Discourse Analysis 70 Conversation

Academic writing

60

Rate per 1,000 words

50

40

30

20

10

iti o as nal ad ph ve ras rb es ia ls

os ep

ify od em Pr

Pr

in g ph pr ra ep se os s itio in n N al Ps

od em Pr

At

tri

bu

te

s

ify

in

ad

g n in ou N ns Ps

je c in tiv N es Ps

0

Figure 9.2 Common dependent phrasal types (from Biber and Gray 2010)

a surrogate for a suite of unmeasured covariates) in academic prose is also particularly salient. Overall, this analysis shows that the stereotype that academic writing is more structurally complex in terms of embedded dependent clauses is not supported by the corpus evidence. While conversation typically employs a higher amount of subordination (embedded dependent clauses) than academic writing, academic writing employs more structures embedded in the noun phrase. These patterns are evident in the following two excerpts. The first text excerpt is from conversation, where the use of subordinate clauses is common and the use of nouns and noun phrase modifiers is less common. In contrast, the second excerpt is from academic prose, illustrating the dense use of modifiers within the noun phrase, including both phrasal structures and relative clauses. In both excerpts, nouns are bolded, noun postmodifiers (finite and non-finite relative clauses, and postmodifying prepositional phrases in noun phrases) are underlined, attributive adjectives and premodifying nouns are in italics. Dependent

148

HylandK_Chap09_Finals.indd 148

12/1/2010 8:04:41 AM

Corpus Approaches to the Study of Discourse

clauses which are sentence elements are double-underlined (if the dependent clause overlaps with other features, only the first few words of the clause are double-underlined). Excerpt 1: From an unscripted organizational board meeting 1: The thing is we only need one funding this year so if they don’t VENRC should fund us 2: Right, but 3: That is true. Or the state department. I talked to Cheryl and, and she thought maybe they could come up with some. 2: Is it that bad a year? 3: Yeah, yeah. They’re eliminating all the jobs except hers at, you know, in Oklahoma. It’s really getting . 2: So, just take it out completely? 3: Um, ... yes, take it out completely. Cause we don’t know yet whether she’ll ... I’m pretty sure. I just sent the, uh, and the bio two days ago to so maybe because she didn’t have that she didn’t want to commit. 1: That’s an idea. 2: I, I got a copy for you. Did you make a copy for yourself. Okay, 3: I thought if I put it on my desk to make a copy I’ll never get it in the mail so I just put it right in the mail for you. 2: Okay. Well I made two copies to send around for anybody who wants to see it and 1: It’s nice. 3: It is a nice . Excerpt 2: From a research article in Biology A common interest in modeling is to make inferences about the effect of the longitudinal measures on the time to event, but not to make inferences about the longitudinal measures or their projected change over time. The longitudinal model is important for accounting for measurement error and defining the individual’s longitudinal trajectory between times of measurement. Because a complex model for the longitudinal measures may be too complicated to estimate in the joint model, a linear mixed model that is parametric with respect to time is most commonly used. More recently, attention has focused on relaxing assumptions on the model for the longitudinal data, but this research has focused mainly on the distributional assumptions of the random effects and has not addressed the shape of the trajectory, often making simple parametric assumptions for complicated patterns of data. If we are only interested

149

HylandK_Chap09_Finals.indd 149

12/1/2010 8:04:41 AM

Bloomsbury Companion to Discourse Analysis

in modeling the longitudinal component, we may often judge these simple models to be inappropriate. Also, when the true relationship between time and the longitudinal biomarker is nonlinear, the effect that assuming a simple linear longitudinal model has on the estimate of the relationship of the marker and the time-to-event outcome is unclear. Both text excerpts are approximately 200 words long. However, the dense use of nouns in the Biology research article is readily evident at a glance. In addition, the high density of nominal modifiers (attributive adjectives, nouns as noun premodifiers, prepositional phrases and relative clauses as postmodifiers) is also prominent. In particular, note that there are often multiple levels of embedding within noun phrases, such as in: inferences about the effect of the longitudinal measures on the time to event where the head noun inferences is modified by one long prepositional phrase, which contains two additional phrases modifying effect (of the longitudinal measures and on the time to event), and a final prepositional phrase (to event) modifying the head noun time. In contrast, the conversation excerpt relies more heavily on pronouns than nouns, and has only one prepositional phrase modifying a head noun (except hers). However, the conversation excerpt does contain many embedded dependent clauses: if they don’t whether she’ll . . . because she didn’t have that . . . thought (that) . . . I’ll never get it in the mail if I put it on my desk so I just put it in the mail for you who wants to see it These corpus findings and the text excerpts above illustrate the distinctive patterns differentiating conversation from academic writing in terms of elaboration. Each register’s reliance on particular features can be functionally related to the situational characteristics of the register. While academic writing is not ‘complex’ in its use of subordination features, it is complex in that it utilizes phrasal embedding as a means of structural elaboration. Although traditional measures of complexity and elaboration typically focus on subordination, phrasal elements like attributive adjectives, nouns as noun premodifiers, appositive noun phrases, and prepositional phrases as noun postmodifiers are also elaborating in the sense that they add optional, extra information. 150

HylandK_Chap09_Finals.indd 150

12/1/2010 8:04:41 AM

Corpus Approaches to the Study of Discourse

However, these phrasal elements are condensed in the sense that they are compressed alternatives to fuller clausal modifiers. In academic prose, writers prefer these compact structures because they are more economical. That is, they allow the expert reader to process a great deal of information quite efficiently. Thus, while both conversation and academic writing are complex and elaborated, they are so in dramatically different ways. This sample study has illustrated how corpus-based discourse studies of language in use can uncover systematic patterns of variation across registers. In addition, it has shown that findings resulting from corpus-based analyses can be counter to many of our intuitions and/or assumptions about language use in a particular register. In the final section, we now turn to how corpus linguistics can be used to investigate language structure beyond the sentence.

Investigating Language Structure beyond the Sentence and Future Directions for Corpus-based Discourse Analysis While most corpus linguistic studies of discourse fit within the language in use framework, it is also possible to investigate text structure, or language beyond the sentence, from a corpus perspective. Research in this area has been limited; one of the few book-length contributions is Biber et al. (2007), which combines the analysis of functional discourse units with corpus linguistic methods, developing a comprehensive framework for applying corpus-based analysis to the analysis of text structure. Biber et al. present two types of discourse units: those derived from top-down analysis (e.g. move analysis) and those derived from a bottom-up corpus-driven analysis (e.g. ‘vocabulary based discourse units’). The top-down approach is a more traditional approach in which each text in the corpus is first segmented into ‘moves’ (see Swales 1990). These moves are analysed in terms of their functions in the discourse, and then analysed for their use of lexical and grammatical features. Once each individual move is analysed in these ways for each text in the corpus, general patterns can be identified for (a) the typical linguistic features of each functional category of moves, and (b) the sequence of moves and functions across the texts in the corpus. For the bottom-up approach, texts are first segmented into discourse units based on linguistic criteria such as shifting vocabulary use. Each discourse unit is then analysed for its linguistic characteristics and classified into linguistically defined categories (rather than functional), and the patterns across the texts of the corpus can be described (see Biber et al. 2007: Ch. 1). In short, applying corpus analysis methods to the study of text structure has the same benefits as corpus analyses of language in use. While each individual discourse unit is analysed first, the end goal is to identify patterns across the texts of the corpus, leading to generalizable observations about the target 151

HylandK_Chap09_Finals.indd 151

12/1/2010 8:04:42 AM

Bloomsbury Companion to Discourse Analysis

register. In addition, because large amounts of data can again be analysed in quantitative and qualitative terms, comparison can be made across registers or genres. This is in contrast to traditional studies of text structure which have focused on detailed, primarily qualitative descriptions of a single text or a very small number of texts of a single genre. Biber et al. (2007) have thus shown that it is useful to use corpus analysis to analyse text structure, integrating quantitative, generalizable analyses with qualitative descriptions. This methodology, while informative, has yet to be widely used to compare text structure across registers, and future studies looking at a variety of registers would certainly add to our knowledge (one exception to this is Upton and Cohen 2009). In addition, other research has integrated corpus-based research with traditional discourse analysis approaches. For example, Garcia (2007) investigates pragmatic utterances in a corpus of university conversations, applying a corpus-based linguistic analysis within Searle’s (1969) pragmatic framework. O’Halloran (2009) and Baker et al. (2008) combine corpus analyses with critical discourse analysis. These studies all reveal the unique contributions that corpus analysis can bring to the field of discourse analysis. As researchers increasingly apply corpus linguistics research in innovative ways to address questions about discourse, we will continue to discover just how much corpus linguistics has to offer.

Note 1. See Lee (2001) and Biber and Conrad (2009, chapter 1) for a comprehensive discussion of the uses of these terms; also see Martin (this volume), Tardy (this volume).

Key Readings Baker, P. (2006), Using Corpora in Discourse Analysis. London: Continuum. Biber, D., Connor, U. and Upton, T. (2007), Discourse on the Move: Using Corpus Analysis to Describe Discourse Structure. Amsterdam: John Benjamins. Biber, D. and Conrad, S. (2009), Register, Genre, and Style. Cambridge: Cambridge University Press. Biber, D., Conrad, S. and Reppen, R. (1998), Corpus Linguistics: Investigating Structure and Use. Cambridge: Cambridge University Press. Charles, M., Pecorari, D. and Hunston, S. (eds) (2010), Academic Writing: At the Interface of Corpus and Discourse. London: Continuum. Flowerdew, L. (2008), Corpus-based Analyses of the Problem-Solution Pattern: A Phraseological Approach. Amsterdam: John Benjamins. McEnery, T., Xiao, R. and Tono, Y. (2006), Corpus-based Language Studies: An Advanced Resource Book. London: Routledge.

152

HylandK_Chap09_Finals.indd 152

12/1/2010 8:04:42 AM

Part II Research Areas and New Directions in Discourse Research

HylandK_Chap10_Finals.indd 153

12/1/2010 8:05:03 AM

HylandK_Chap10_Finals.indd 154

12/1/2010 8:05:04 AM

10

Spoken Discourse Joan Cutting

Chapter Overview Types of Spoken Discourse Spoken Discourse Features Function of Features Social Variables Sample Study Conclusion Transcription Key Key Readings

155 158 162 165 166 169 170 170

With the advent of sensitive digital recording equipment, and software to transfer spoken discourse to a computer and enable systematic studies, linguists’ understanding of spoken language is now catching up with that of written language. This chapter provides an introduction to types of spoken discourse and their characteristics, and discusses the functions and social variables that influence the choice of spoken discourse features.

Types of Spoken Discourse It is commonly understood that spoken language is a reflection of the process of language construction whereas written language is a revised and polished product. However, the difference is more subtle than that. There is a cline from spontaneous spoken discourse (unplanned and semi-planned) to non-spontaneous spoken (semi-scripted or scripted) to spontaneous written discourse (unplanned and semi-planned) to non-spontaneous written discourse (polished scripts), and the borders between one category and the next are blurred. Most spoken discourse is unplanned: speakers do not know exactly what they are going to say before they say it, and they put their words together as 155

HylandK_Chap10_Finals.indd 155

12/1/2010 8:05:04 AM

Bloomsbury Companion to Discourse Analysis

they are talking. The prototypical unplanned spoken discourse is the casual conversation, examples being gossip between friends over coffee, chat between strangers at a bus stop, and informal health-reporting between a mother and daughter on the phone. The unplanned nature of an exchange can make topics shift seamlessly; note how in excerpt (1), from a conversation between fellow students in a university common room, the interactivity contributes to the unpredictability of its direction. Excerpt (1) AF Also. I had quite a bad cold last week and (3) I didn’t feel much like going out (1) DM Feel better now? AF Mm. Yes the worst was actually in the middle of the week when I was planning to work very hard. Cos em I’d got it over with when I was out without the opportunity to succumb as it were. So (heh heh) (8) I suppose we were out and about on Saturday afternoon. Actually I was up at the City Cafe. DM Mm? AF Heaping plates of chips into Julian. They do very good chips there actually. DM Where’s that? AF The City Cafe on Blair Street. (2.5) Just above Cowgate. You know when we go down to Wilkie House. DM Aha. AF Well the street we go down the City Cafe is in that. It’s a nice place. (3) And it’s like the only drinking place in Edinburgh where you can take children. DM And get chips. AF Just so. DM Perfect. Conversation takes place in a shared context; it is interactive, interpersonal and informal, and contains expressions of wishes, feelings, attitudes and judgements (Thornbury and Slade 2006: 8–25). In excerpt (1), AF and DM share knowledge: AF does not have to explain what Wilkie House is or who Julian is. They show feelings (‘the worst was actually in the middle of the week’) and make judgements (‘They do very good chips there actually’, ‘Perfect’). Much spoken discourse is semi-planned, in that the speakers have an idea about the sort of thing that they are going to say before they say it. In this category are difficult conversations that need carefully chosen rehearsed words, such as a parent telling a pubescent child about making babies, and a man 156

HylandK_Chap10_Finals.indd 156

12/1/2010 8:05:04 AM

Spoken Discourse

proposing marriage to his sweetheart. Others in this category are job interviews, interactive learner-centred classes and some public speeches. Excerpt (2) is from a lecture from The British Academic Spoken English (BASE) corpus. The fact that it ‘trips off his tongue’ suggests that the lecturer had planned, if not rehearsed, his words: Because what we are grappling with in the phenomenon of imperialism is a phenomenon that in various forms is as old as the formation of state systems by human beings. So I’m going to, er, at considerable risk er to myself, try to set this phenomenon in a much wider, er, more global perspective. I hope that might be of interest to many of you who have either been subjected to what you consider imperialism, or indeed have been part of states and societies that have themselves been imperialistic or are still being so. The less interactive the discourse is, the more control the speakers have over the topic development and direction, and the words used. Semi-scripted spoken discourse is similar to semi-planned except that some of the planned words have been written down. It differs from scripted in that it allows for creativity. Semi-scripted discourse occurs most frequently in encounters where the speakers are trained to use basic routines and short formulae, where similar utterances are repeated with each hearer, or where the speaker is guided by notes. Witness the shop assistant’s ‘Pop in your pin-number and press enter’, the hairdresser’s ‘Going anywhere nice for your holidays?’, or the pilot’s ‘Sit back and relax and enjoy the flight’. If a university lecturer reads from the PowerPoint slides, or a public speaker uses a written list of points, it is semi-scripted. Scripted spoken discourse is on the borderline with written discourse because it amounts to reading aloud written words. In this category fall recorded telephone messages, news reports on radio and TV, plays and films, and lectures and speeches that are read out. These class as spoken discourse as they are intended to be heard, and because readers can add meaning and feeling through stress and intonation, and there is disfluency if they mispronounce words, stumble and repeat words, or cough and apologize mid-sentence. Spontaneous written discourse, on the borderline with spoken, can be unplanned or semi-planned writing such as scribbled notes on the back of an envelope, mind maps for an essay, text messages, emails, chat-room text, blogs and Facebook messages, which tend to be informal, and often interactional and involved interpersonally. Cameron (2001: 9) points out that computer-mediated communication shows that interactivity and spontaneity are ‘more salient characteristics of “conversation” than the channel or medium of interaction’. Crystal (2004: 69–76) argues however that netspeak is different from face-toface conversation in that there is a lack of simultaneous feedback, a possibility of multiple simultaneous conversations and a slower rhythm of interaction. 157

HylandK_Chap10_Finals.indd 157

12/1/2010 8:05:05 AM

Bloomsbury Companion to Discourse Analysis

Spoken Discourse Features Describing the features of spoken discourse has its problems. Its tendency away from standards and collective norms, its personal character, its layers of meaning and function, and its fast-changing nature, make it relatively difficult to write a grammar book to describe it in a way that all speakers can agree with. Another problem is that there is not one standard terminology: Biber et al.’s (1999) is slightly different from Carter and McCarthy’s (2006), which in turn is slightly different from Huddleston and Pullum’s (2002). To take an example of variance in the naming of parts, expressions such as ‘or something’ and ‘and all the rest of it’ are labelled ‘vague category identifiers’ by Channell (1994), ‘general extenders’ by Cutting (2007) and ‘tags’ by Thornbury and Slade (2006). The following description of the rules of spoken grammar, in the wider sense of lexis, syntax and disfluency, is taken from McCarthy (1999), Cutting (2000), Carter and McCarthy (2006) and Thornbury and Slade (2006). The lexical features are: z Vague words

nouns: superordinate nouns, empty-semantically, dependent ◦ general on the context for meaning, for example ‘I brought that thing’, ‘I’ll never

◦ ◦ ◦ ◦

get this stuff’, ‘work with one other person’, ‘crummy places that I’ve been working for’, ‘I haven’t got my thingymajig’ general noun clusters: fixed vague expressions, for example ‘I brought the what’s-a-name’, ‘ask what’s-her-face’ general verbs: superordinate verbs (‘do’, ‘happen’), empty-semantically, dependent on the context for meaning, for example ‘I’ve done my thing’, ‘They did all the stuff’ general extenders: vague endings, referring to the rest of a category or just a filler, for example ‘With a little heading here or something’, ‘Look it up under insert table or something like that’, ‘hedging and surveying and so on’, ‘You don’t have to sort of phone and make arrangements, and things’ vague quantifiers: non-specific expressions of quantity, for example ‘She’s got lots of things to tell you’, ‘They’ve got loads of money’

z Informal words

names, nicknames and informal forms of address: for example ◦ short ‘we got to one that Dave and I usually go to’, ‘Morning, Tootsie!’, ‘You

◦ ◦

get me, blood?’ delexical verbs: common verbs (‘give’, ‘have’, ‘make’) that combine with nouns to describe an action, for example ‘take a break’, ‘have a bath’ basic adjectives and adverbs: high-frequency, everyday words, for example ‘It’s a nice place’, ‘Are you shutting out this lovely sunshine?’ ‘Basically we’ve finished’, ‘Anyway eventually after walking round the pubs, we gave up’

158

HylandK_Chap10_Finals.indd 158

12/1/2010 8:05:05 AM

Spoken Discourse

colloquial and swear words: ‘pretty warm’, ‘You’re an absolute dosser!’, ◦ ‘Bloody hell – is it eleven already?’ z Interactional word chunks

markers: words marking boundaries between topics and ◦ discourse stages; opening, closing, sequencing, for example ‘Right, let’s get

◦

◦ ◦

started’, ‘Anyway I decided to, so that was that’. fillers and hedges: words and sounds filling pauses, marking hesitations, delaying words, for example ‘um – er, OK’, ‘Like I didn’t even know what he wanted’, ‘It’s a bit abstract’, ‘Kind of scary actually’, ‘lots of you know sort of pushing and shovelling around’, ‘But at the end of the day, it doesn’t matter’, ‘like I didn’t even know what he wanted’ communication checks, confirmations and backchannels: expressions that keep the channel of communication open, for example ‘She’s just lazy, do you know what I mean?’, ‘Oh I see yeah’, ‘Mhm, aha’ expressions of politeness: ‘Yeah thanks’, ‘Could you get me a tuna and sweet corn one please?’

The syntactical ones include: z Vague reference

pronouns: for example ‘she’ll look for something else’, ◦ indefinite ‘nobody’s going’ determiners (‘this’, ‘those’, ‘here’, ‘then’) and personal pronouns ◦ deixis: (‘I’, ‘we’, ‘him’, ‘us’) pointing to a referent in the situational or background

◦

context, for example ‘That was yesterday’, ‘Probably a bit late now’ general noun clauses: vague expressions, not fixed, empty-semantically, for example ‘Are you going to do what you thought you’d do?’,

z Incomplete utterances

clausal ellipsis: omission of subject and or verb, for example ◦ initial ‘Perfect’, ‘Been there. Done that’ subordinate clauses: subordinate clause expressed as a sep◦ stand-alone arate utterance, for example ‘We went to the Tron. Which is that square’ utterances: omission of the end of the utterance, for exam◦ unfinished ple ‘And I knew I was going to do one essay so . . . ’

z Informal grammar

clauses: joined by ‘and’, ‘but’, ‘then’ and ‘because’. ◦ short headers: adjuncts, objects and complements, and noun phrases ◦ before thefronted pronoun, for example ‘The film I liked’, ‘John, the man I lived

◦

with, the Scot, he played blues guitar’ tails: noun or prepositional phrases cohering with the pronoun, after the clause, for example ‘They’re slow, these traffic lights.’, ‘It’s lovely, Blackpool is.’ 159

HylandK_Chap10_Finals.indd 159

12/1/2010 8:05:05 AM

Bloomsbury Companion to Discourse Analysis

grammar: systematically used spoken grammar that breaks ◦ vernacular written grammar rules, for example ‘He vowed to boldly go’, ‘Give it to Sue and I’, ‘He wanted to know why was I there’, ‘They usually always eat out on a Sunday’, ‘I didn’t do nuffink, like, innit’ The disfluency features are: z repetitions – for example ‘She just couldn’t remember if we were on the –

the right track.’ z recasting – repairs and reformulations mid-utterance, for example

‘Actually – in it – and one guy – there was three guys on the stage.’ z pauses – unfilled hesitations of half a second or more, for example ‘It was

in em – where was it (1) in em in em (0.5) George Street.’ z overlaps – (often indicated with =) simultaneous speech when the first

speaker completes their utterance, for example Speaker A: ‘Olivia Newton John and em = whatever the character’s called. John Travolta.’ Speaker B ‘= (heh heh) Hang on. Where did you see this?’ z interruptions: (often indicated with //), simultaneous speech when the first speaker does not complete their utterance, for example Speaker A: ‘Um. Well what happened is // that’ Speaker B: ‘// I mean when you wake up are you sitting on the bench?’ These features were initially thought to be typical of only unplanned speech, but in fact they occur in all spoken discourse. It would be safe to say that the closer to the unplanned end of the cline, the greater the density of these features. In unplanned excerpt (1), there are colloquial words (‘Cos em I’d got it over with . . .’), fillers (‘Actually I was up at the City Café’, ‘Well the street we go down the City Cafe is in that’), initial clause ellipsis (‘Feel better now?’) and headers (‘Well the street we go down the City Cafe is in that’). In semi-scripted excerpt (2) we just meet a stand-alone subordinate clause (‘Because what we are grappling with in the phenomenon of imperialism is a phenomenon that in various forms is as old as the formation of state systems by human beings.’) and fillers (‘So I’m going to, er, at considerable risk er to myself, try to set this phenomenon in a much wider, er, more global perspective’). Spoken discourse is said to have lower lexical density and less intricate grammar than written discourse. It is indeed usually the case that it has fewer content words (nouns, verbs, adjectives, adverbs) than function words (grammatical words such as prepositions, pronouns and articles), but it can often have embedded clauses and long complicated sentences. Witness this student’s description of her essay-writing process, with its short clauses joined by ‘and then’, ‘but also’ and ‘and then’: 160

HylandK_Chap10_Finals.indd 160

12/1/2010 8:05:05 AM

Spoken Discourse

Excerpt (3) BM Last night I was trying to I had some sort of about page four I had I’d written a whole page with normal paragraphs and then I realised I’d indented the whole thing in and but also at the same time I wanted on the left hand margin on the left hand margin I wanted em a number. (0.5) and then some information between the number and the text which was indented = CM = It has to do It’s called = BM = It’s called a bloody (0.5) disaster. (0.5) But it’s a I had I went through and I em of course indented every line and cycled and cycled the thing. Spoken discourse tends to be more implicit and heavily dependent on the context for its meaning than written. In excerpt (4), the speakers use general verbs and nouns and an indefinite pronoun to refer vaguely to specific actions and referents that they both know. Excerpt (4) DM So did you do everything? DM // Or sort of choose a few people? CM // Em. (0.5) CM No I’ve done all the people. DM // Really? The clustering of vague features like this can exclude outsiders. In excerpt (5), the implicitness is at a deeper, utterance level: there is a whole implied subtext in last three discourse units dependent on interpersonal knowledge of family situation and of what is acceptable humour: Excerpt (5) AF Oh I’ve got no money. AF I’ve got to go to the bank. DM same here. BF No money? BF What do you mean you’ve got no money? (2) BF You’ve a working wife. DM That’s why I have no money. On the other hand, implicit language is not exclusive to spoken discourse. Take written advertisements for example; they often use implicitness to influence consumers. Readers in Britain will have seen posters for a Do-It-Yourself chain claiming ‘You can do it when you B&Q it’, the general verb and exophoric personal pronoun covering a multitude of DIY. A written note on the refrigerator door that says, ‘Your dinner’s in the dog’ tells the reader implicitly that their partner is angry that they came home so late. 161

HylandK_Chap10_Finals.indd 161

12/1/2010 8:05:06 AM

Bloomsbury Companion to Discourse Analysis

It goes without saying that not all unplanned speech is vague and informal. A spur-of-the-moment ‘Let’s meet in the foyer of the Film House at 7.45, get our tickets and go for a quick coffee in the bar’ needs to be that precise and explicit. An unexpected encounter’s ‘Dr Xiao, I’d like to introduce you to my head of department, Professor Webber’ demands a certain formality.

Function of Features The function of spoken discourse has been studied using a variety of approaches. Critical Discourse Analysis (CDA) (Fairclough 2003) is widely used (see Wodak’s chapter in this book). A tried and tested approach is Speech Act (SA) theory (Austin 1962, Searle 1969), which looks at the functions of language, for example promising, offering, threatening, inviting, suggesting, ordering, apologizing, praising and regretting. SA theory answers questions like ‘Do advanced learners of X language use all the range of vague expressions that native speakers do when apologizing?’ and ‘How do people promise, offer and threaten indirectly?’ Another approach is the Cooperative Principle (CP) (Grice 1975), which explains how speakers keep conversations going smoothly, following maxims about being sincere, giving the right amount of information, being relevant and clear. CP answers questions such as ‘Are certain disfluency features associated with lying?’ and ‘Are the four maxims flouted in the same way across cultures X, Y and Z?’ The last approach to be mentioned here is the Politeness Principle (PP) (Brown and Levinson 1987), which is about saving face, by being indirect, not imposing and keeping respectful distance, or showing closeness and solidarity. PP answers questions such as ‘Do speakers avoid offending by using vague reference and incomplete utterances?’ and ‘Do some cultures put more value on showing respectful distance than other cultures?’ To demonstrate these approaches, let us examine some published studies. Speech acts in spoken corpora have been analysed by many. Adolphs (2008: 43–88) focused on suggestions in the Cambridge and Nottingham corpus of Discourse in English (CANCODE) and the variables of relationship, domain and topic. She discovered that the informal expression ‘why don’t you’, as in ‘Why don’t you just wait?’ is used more between intimates, whereas the verb ‘suggest’, as in the more formal ‘I wish to suggest simply that . . . ’, is used more in pedagogic exchanges. She found that the ellipted ‘how about’, as in ‘How about finding out?’ was used in discussion of personal problems. Many studies have investigated the functions of spoken discourse features using CP and PP. Let us take the example of vague language, which ‘nearly always enables polite and non-threatening interaction’ (Carter in Seidlhofer 2003: 92). Jucker, Smith and Lüdge (2003) found that vague quantifiers soften complaints and criticisms and establish a social bond. Ruziaté (2007: 213) 162

HylandK_Chap10_Finals.indd 162

12/1/2010 8:05:06 AM

Spoken Discourse

discovered that vague approximators were used by students and teachers in spoken academic discourse, ‘to shield their claims against possible criticism, avoid categorical claims, observe the politeness principle and save face’. Drave (2002) suggested that vague language promotes politeness and closeness, and is used for managing asymmetries of knowledge in intercultural interaction. Cutting (2000) discovered that an academic discourse community’s use of implicit features constituted a high involvement strategy for asserting ingroupness. Cutting (1998) also found that the most frequent function of vague language in the CANCODE was that of giving little importance to the referent, either to be friendly or to be critical. Different approaches can give different interpretations of what is going on in any one stretch of spoken discourse. An analysis of excerpt (6) about Edinburgh buses (Cutting 2000) will demonstrate this. Excerpt (6) BM I had a really bizarre conversation with a bus driver over that and em it was early in the morning and he had the particular kind of accent that I c- didn’t quite catch everything he said. He didn’t quite catch everything I said either. So it was working both ways. So we had this really confusing conversation. (0.5) I was on the wrong bus. But anyway I’d got on it. BF (heh heh) (1) We had a really neurotic bus driver this morning. Cos he (0.5) there must be something wrong with his hearing and everybody was getting on and saying forty please. He said well what kind of fare do you want? You know (heh heh heh) DM // (heh heh heh) BM // (heh heh) BF I wish you’d speak up you know. (heh heh) Really annoyed with everybody. DM Demented bus driver. BF // (heh heh heh) AF // (heh heh heh) You get the occasional one. (2.5) The wild ones are on the on the green buses actually. They go at a hell of a rate. Don’t stop for anything. If you got to // want to get somewhere BF // Limited stop buses are they? AF No just the no they’re just em the the SMT buses // the green ones that go BF // Oh. AF into Saint Andrew’s Square. The big ones and the little ones and the (0.5) I d- I don’t know. I think they’re on some sort of productivity bonus or something. BF (heh heh) 163

HylandK_Chap10_Finals.indd 163

12/1/2010 8:05:06 AM

Bloomsbury Companion to Discourse Analysis

DM AF BF DM AF BM AF BM DM BF DM AF

What the more times they go round. // (heh heh) // Yeah something like that. Sponsored bus drivers. // (heh heh) // (heh heh heh) // (heh) They give change don’t they. They do. They’re sort of we give change! If you can get on the bus you get change then. Ah right. // That’s good. // Yeah. If you don’t mind breaking your leg as you try and get on. (heh heh)

Taking a PP approach, one could reach the conclusion that the abundance of spoken discourse features suggests that all interlocutors wish to cohere socially and show solidarity by adopting the same relaxed style. They all use general extenders (‘I think they’re on some sort of productivity bonus or something’), discourse markers (‘But anyway I’d got on it’), fillers (‘You know’), communication checks and confirmations (‘Ah right’) and indefinite pronouns (‘Don’t stop for anything’). The initial clausal ellipsis (‘Demented bus driver’, ‘Sponsored bus drivers’), stand-alone subordinate clauses (‘If you don’t mind breaking your leg as you try and get on’) and unfinished utterances (‘What the more times they go round’) make the interaction sound like one shared utterance. PP analysis would reveal that the speakers are showing positive politeness by claiming common attitudes and knowledge. When BM says that he ‘had a really bizarre conversation with a bus driver’, BF laughs to encourage him and emphasizes her common attitude by telling a similar story about a funny driver, and the others laugh empathetically, DM joining in the fun by evaluating BF’s driver the same way. All four of them then exaggerate the speed of the buses, for their mutual amusement. BM adds a new dimension with his ‘They’re sort of we give change!’, DM shows solidarity by echoing the words and augmenting the story, and AF takes this one stage further again, linking the change joke to the speed joke. What emerges is a story-telling routine, a cosy bonding activity, in which speakers entertain each other. If on the other hand, the conversation is analysed using CDA, a different picture emerges. We could notice that the speakers have concealed purposes and biases. BM asserts his assumption that laughing at a bus driver who does not hear well is acceptable. This persuades BF to enter into this politically incorrect game: she imposes her assumption that it is all right to laugh at people who seem ‘neurotic’ and this makes DM portray ‘demented’ people in the same vein. Once disabilities and mental illness are established as entertaining, nobody dare deny it or spoil the fun. 164

HylandK_Chap10_Finals.indd 164

12/1/2010 8:05:06 AM

Spoken Discourse

Power is expressed in language, and knowledge is power. AF as an Edinburgh resident has power over the in-comers: she talks about ‘the green buses’ without explaining, and when BF interrupts her with a request for an explanation, she carries on with further Edinburgh terms that do not help her. BF gives up with a single ‘Oh’; AF maintains her power. BM then adds what seems an irrelevant point about giving change; as the other Edinburgh resident, and possibly as a male not wanting to be out-done by a female, he needs to show that he too has knowledge of Edinburgh. AF does not encourage him to take over the centre stage position, with her minimal ‘They do’. DM, the other male, then seizes the chance to take over. CDA would see this as a power struggle, with the interlocutors competing to see who is the most knowledgeable, the best story-teller and the nicest to be with. It is uncertain which of these two interpretations is ‘right’. It could be that they both are, or that neither is. Asking the people recorded often does not help in interpretation, as intentions can go on at multiple subliminal levels. All the researcher can do is guess.

Social Variables Spoken discourse needs to be seen in its social context: the language that any one person uses varies according to who they are talking to, where they are and what their intention is. It varies according to the country it is spoken in, the discourse community it belongs to, and the age, gender, class and educational level of speakers. Compare the Scottish statement with its double header ‘See my man, see mince, my man likes mince’, with the Standard English ‘My man, he likes mince.’ Note how educated people in Britain might ask ‘After all, he’s good, isn’t he?’, whereas less educated people might use the vernacular ‘At the end of the day, he’s ace, innit?’ Corpus linguistics and sociolinguistics are coming together but there has been a lack of sociolinguistic metadata encoded in corpora, and a lack of sociolinguistically rigorous sampling in corpus construction (McEnery and Wilson 2001). Spoken corpora have however facilitated sociolinguistic studies. For example, the Scottish Corpus of Texts and Speech allows users to narrow their word-search down according to gender, region, mode and year composed, and the Michigan Corpus of Academic Spoken English permits users to call up lines with variables of gender, age-range, academic role and first language. Later chapters in this book cover social variables. Let us here take two examples of gender and language studies, as relates to disfluency and interaction. Coates (1993) found that men interrupt women more than they interrupt other men, far more than women interrupt men and more than women interrupt

165

HylandK_Chap10_Finals.indd 165

12/1/2010 8:05:06 AM

Bloomsbury Companion to Discourse Analysis

other women. Jenkins and Cheshire (1990) discovered that women give more backchannel support than men do, as in ‘mmm’, ‘uhuh’, nodding and smiling.

Sample Study The mini-study (Cutting 2009) described for this chapter is a pilot for part of a larger project about the spoken interaction of international students in UK universities. It examined video-recordings of five tutorials in TESOL Masters degree courses. Table 10.1 shows the tutorial composition. Although the tutorials were all about three hours long, only the parts which were not in lecture mode were analysed. The research question was ‘Do lecturer linguistic features correlate with student talking time?’ The academic discourse community in-group code outlined by Cutting (2000) was used as the linguistic features model. Thus the lexis included technical terms (e.g. ‘task-based learning’, ‘schemata’), general nouns (e.g. ‘they can recognize the whole thing’), general verbs (e.g. ‘what happened?’), colloquial words (e.g. ‘This is a bit dodgy for health and safety’) and short names, and the grammar consisted of deixis (e.g. ‘That’s not what they advertised’, ‘Whose is this?’), indefinite pronouns (e.g. ‘Everybody ready?’), discourse markers (e.g. ‘Right, let’s move on then’), fillers and hedges (e.g. ‘They can sort of work out’, ‘we try er very hard’) and backchannels (e.g. ‘Aha aha’, ‘Mhm’). The fact that in this project, some syntactical, interactional and disfluency features were considered to be spoken grammar, and other disfluency features (overlaps and interruptions) were called structural features, shows that there is no one way to conceive and arrange these elements. The density of lecturer linguistic features was measured as the percentage of features out of all lecturer words. Student talking time was measured in seconds.

Table 10.1 Tutorial length and content Tutorial Duration

Words

Content

1hr 30mins

4,076

2hrs 47mins

8,250

2hrs

8,825

2hrs 30mins

16,303

20mins

1,040

student presentations, plenary discussion lecturer interactive lecture, plenary discussion student presentations, plenary discussion student posters, small-group discussion student presentations, plenary discussion

166

HylandK_Chap10_Finals.indd 166

12/1/2010 8:05:06 AM

Spoken Discourse

Quantitative analysis showed differences between features and tutorials but since the study had such a small database, the findings are not generalizable. The lecturers seemed to use more of the grammatical features than the lexical (see Table 10.2). This is hardly surprising as pronouns and interactional word chunks are more frequent than technical terms and vague informal lexis in most spoken interaction, because of the low lexical density in semi-planned spoken discourse. This result is possibly also a reflection of the decision to include interactional word chunks in the ‘grammar’ category. There was a higher density of lecturer linguistic features in tutorial 1 than in tutorials 3 and 4. This might be because tutorial 1 lecturer had a more relaxed manner than the others. Tutorial 1 had less student talking time than tutorials 3 and 4 (see Table 10.3). In the case of tutorial 4, it could be that the poster and small group discussions gave students more opportunities to talk. However tutorial 3 had the same format as tutorial 1, so it is not clear whether the pedagogical technique influenced the amount of student talking. It appears that the more lecturers used the in-group code features under examination, the less the students participated. A more extensive study could test whether this negative correlation is a widespread tendency. A qualitative impression of the data seemed to support this new hypothesis. When there was a clustering of the lecturer linguistic features, student participation appeared to be hampered. In excerpt (7), it could be that the lecturer’s indefinite pronoun, technical word and colloquial verb, possibly unknown

Table 10.2 Percentage of lecturer linguistic features in each tutorial 14

Lexical Grammatical

% features/all wds

12 10 8 6 4 2 0 1

2

3 tutorials

4

5

167

HylandK_Chap10_Finals.indd 167

12/1/2010 8:05:07 AM

Bloomsbury Companion to Discourse Analysis

Table 10.3 Percentage of lecturer and student talking time in each tutorial 100 90 80 % in seconds

70 60 Lecturer

50

Students

40 30 20 10 0 1

2

3

4

5

to the non-English-native-speaker students, explain the low student talking time. Excerpt (7) LF3 So, can you give me something under the heading of micro-skills? (3) What was it? Eh, you two have a very good satisfying agreement. What would you put your success down to? (2) How did you start your conversation? S4 Eh, self-introduction. LF3 Aha S4 And eh LF3 OK. Note the long pauses when the lecturer is waiting for the students to answer, and the fact that she has to re-phrase her question three times; the student’s responses are almost mono-syllabic. In excerpt (8) it may be that the lecturer’s general nouns and verbs are responsible for the short student turns Excerpt (8) LF2 = What sort of things that really actually happen (0.5) with eh, adjectives, when do people use a lot of adjectives? = S3 = Mm = LF2 = You know, you have to try and think // S3 // Descriptions 168

HylandK_Chap10_Finals.indd 168

12/1/2010 8:05:07 AM

Spoken Discourse

LF2 S2 LF2 S2 LF2 S3 LF2 S2

Yeah. Yes. But but when will you use adjectives? Relevant contexts What kind of, sorry? Relevant contexts. Yeah. What what would they be, you know? (explains to OM2) She means in real life. Yeah. What would // // Describe people.

The lecturer’s recasting, pause, filler and repetitions suggest that she might be struggling to make herself understood. In the third example from this study, the lecturer’s question contains a general noun, a filler, a pause and repetitions: Excerpt (9) LM1 Eh, are those (0.5) are those texts going to have all of those activities or are those activities going to run across, some with one text and some with another text, and or what, what’s your thinking there? (3) S1 Eh. (6) LM1 It’s OK to say “ I haven’t thought about that so far.” // OK? All // (laugh) It is, in all probability, not the general noun or disfluency features that flummox the students however, but the complexity of the question’s grammar, with its ‘or’, its ‘and or’, and its three-questions-in-one structure. The nine-second pause is hardly broken by the student’s ‘Eh’. In summary, while it is undeniable that certain teaching formats, strategies and techniques increase student talking time, it could be that lecturer vague language explains the student silences. Further research might show that other ways of measuring student participation give different results, or that other variables such as lecturer speech acts, PP or CP maxim observance, influence student participation more than lecturer language does.

Conclusion New directions in this field are constantly emerging. Corpus linguistics is likely to incorporate more the other approaches to language analysis and look at language in its socio-functional context. Future studies are likely to explore further the multimodal aspects of spoken discourse, following on after studies of linguists such as Adolphs and Carter (2007) who video-recorded lecturers in tutorials, electronically tracing their head and hand movements, so as to see how active listenership is demonstrated by head-nods matched with verbal 169

HylandK_Chap10_Finals.indd 169

12/1/2010 8:05:07 AM

Bloomsbury Companion to Discourse Analysis

backchannels. With the development of electronic means of communication, there will be a growing number of studies of CMC conversations, and the comparison with face-to-face conversations will show clearer the similarities and differences. In addition, there will be a growth in spoken discourse studies of languages other than English. The increasing number of spoken corpora means that language learning coursebook writers can incorporate naturally occurring data into their materials, and use the findings from studies in their task design. There are books that do this already: Exploring Spoken English (Carter and McCarthy 1997: 7) and Touchstone (McCarthy et al. 2005). However, caution is recommended. Some learners do not want to sound native-speaker-like; others have a negative attitude to the target language culture; for some, vague informal language has negative associations in their mother tongue. As Beebe (1988: 63) pointed out, second language learners ‘may find that the reward of being fluent in the target language is not worth the cost in lost identification and solidarity with their own native language group’. EFL teachers may prefer just to raise awareness, explain native-speaker social associations and provide choices, so that learners can opt in or opt out (Cutting 2005: 174).

Transcription Key = indicates overlap, // indicates interruption, (3), (8) etc. indicate the number of seconds’ pause

Key Readings Adolphs, S. (2008), Corpus and Context: Investigating Pragmatic Functions in Spoken Language. Amsterdam: John Benjamins. Anderson, W. and Corbett, J. (2009), Exploring English with Online Corpora. Basingstoke: Palgrave Macmillan. Cameron, D. (2001), Working with Spoken Discourse. London: Sage. Carter, R. and McCarthy, M. (1997), Exploring Spoken English. Cambridge: Cambridge University Press. Cutting, J. (ed.) (2007), Vague Language Explored. Basingstoke: Palgrave. McCarthy, M. (1998), Spoken Language and Applied Linguistics. Cambridge: Cambridge University Press. Thornbury, S. and Slade, D. (2006), Conversation: From Description to Pedagogy. Cambridge: Cambridge University Press.

170

HylandK_Chap10_Finals.indd 170

12/1/2010 8:05:08 AM

11

Academic Discourse Ken Hyland

Chapter Overview What Is Academic Discourse? Why Is Academic Discourse Important? How Is Academic Discourse Studied? What Do We Know about Academic Discourse? A Sample Study: Citations Conclusion Key Readings

171 172 174 176 182 184 184

What Is Academic Discourse? Academic discourse refers to the ways of thinking and using language that exist in the academy. Its significance, in large part, lies in the fact that complex social activities like educating students, demonstrating learning, disseminating ideas and constructing knowledge, rely on language to accomplish. Textbooks, essays, conference presentations, dissertations, lectures and research articles are central to the academic enterprise and are the very stuff of education and knowledge creation. But academic discourse does more than enable universities to get on with the business of teaching and research. It simultaneously constructs the social roles and relationships which create academics and students and which sustain the universities, the disciplines, and the creation of knowledge itself. Individuals use language to write, frame problems and understand issues in ways specific to particular social groups and in doing these things they form social realities, personal identities and professional institutions. Discourse is at the heart of the academic enterprise; it is the way that individuals collaborate and compete with others, to create knowledge, to educate neophytes, to reveal learning and define academic allegiances. The academy cannot be separated from its

171

HylandK_Chap11_Finals.indd 171

12/1/2010 8:05:29 AM

Bloomsbury Companion to Discourse Analysis

discourses and could not exist without them. No new discovery, insight, invention or understanding has any significance until it is made available to others and no university or individual will receive credit for it until it has seen the light of day through publication. At one level then, the study of academic discourse is interesting for what it can tell us about the accomplishment of academic life. But beyond the university, the languages of the academy have quietly begun to insert themselves into every cranny of our lives, colonizing the discourses of technocracy, bureaucracy, entertainment and advertising. Academic discourses have reshaped our entire world view, becoming the dominant mode for interpreting reality and our own existence. We find traces of it not just in popular science periodicals but in the Sunday broadsheets and the TV documentary, it is the language of the pharmaceutical bottle and the toothpaste advertisement, the psychotherapist and the recycling leaflet. It is the carrier of expertise and prestige – the badge of those who possess knowledge and of those who wish to. As Halliday and Martin (1993: 11) put it: ‘the language of science has become the language of literacy.’ There are therefore good reasons for taking academic discourse seriously, and in this chapter I will seek to show why academic discourse is important, something of what is known about it, and how it is studied, finishing up with a sample study which illustrates these issues.

Why Is Academic Discourse Important? The current interest in academic discourse, and particularly academic writing in English, is largely the result of three major developments over the past 20 years: changes in higher education which have resulted in greater interest given to the importance of writing; the growth of English as the international language of research and scholarship; and the emergence of theoretical perspectives which recognize the centrality of academic discourses in the construction of knowledge. First, many countries in Europe, Asia and Australasia have witnessed a huge expansion of Higher Education as a result of greater social inclusion policies. This expansion has been accompanied by increases in full fee paying international students to compensate for cuts in government support, and by the rapid rise in refugee populations around the world with a consequent increase in international migration. Together these factors have a created a student body which is far more culturally, socially and linguistically heterogeneous than ever before. Added to this is the fact that students now take a broader and more eclectic mix of subjects. The ‘academicization’ of practice-based disciplines such as nursing, social work and marketing and the growth of modular and interdisciplinary degrees means that students have to learn rapidly to 172

HylandK_Chap11_Finals.indd 172

12/1/2010 8:05:30 AM

Academic Discourse

negotiate a complex web of disciplinary specific text-types, assessment tasks and presentational modes (both face-to-face and online) in order to graduate. So while writing continues to be the way in which students both consolidate and demonstrate their understanding of their subjects and are socialized into academic practices, students, including native English speakers, must take on new roles and engage with knowledge in new ways when they enter university. An important result of these changes is that learners bring different identities, understandings and habits of meaning-making to a more diverse range of subjects, so that tutors cannot assume their students will possess the understandings and learning experiences that will equip them with the literacy competencies traditionally required in university courses. As a result, greater emphasis is now placed on academic literacy and on EAP programmes to help students meet the demands of their courses. Academic discourses have been extensively studied to inform this pedagogic agenda. A second reason for this growing interest in academic discourse has been the power it wields in the careers of individual academics. Publishing is the main means by which academics establish their claims for competence and climb the professional ladder. Moreover, as pressures on academics to publish increases, so does the demand that this should be done in English. Research shows that academics all over the world are increasingly less likely to publish in their own languages and to find their English language publications cited more often. There were over 1.1 million peer-reviewed research articles published globally in English in 2005 and this number has been increasing by 4 per cent annually. With publishers encouraging libraries to subscribe to online versions of journals, the impact of English becomes self-perpetuating, since it is in these journals where authors will be most visible on the world stage and receive the most credit. This has meant that the numbers of non-native English speaking academics publishing in English language journals now exceeds papers authored by native English speakers (Swales 2004), driving a demand for writing for publication courses. In this enterprise the study of academic discourse has become central to pedagogy. A third major incentive for studying academic discourse comes from a very different direction: the questioning of a positivist, empirical view of scientific knowledge. In recent years the view of academic discourse as an objective, independent demonstration of absolute truth has been challenged by the sociology of scientific knowledge. Essentially, this perspective argues that scientific proof does not lay in the application of impartial methodologies but in academic arguments. Observations are as fallible as the theories they presuppose, and so texts cannot be seen as accurate representations of ‘what the world is really like’ because this representation is always filtered through acts of selection and foregrounding. In other words, there is always more than one possible interpretation of data and these competing explanations shifts attention from 173

HylandK_Chap11_Finals.indd 173

12/1/2010 8:05:30 AM

Bloomsbury Companion to Discourse Analysis

the laboratory or clipboard to the ways that academics argue their claims. We have to look for proof in the textual practices for producing agreement. At the heart of academic persuasion, then, is writers’ attempts to anticipate and head off possible negative reactions to their claims and to do this they use the discourses of their disciplines. Interest in academic discourse has therefore emerged as part of attempts to reveal the specific rhetorical practices of academic persuasion. Analysts seek to discover how people use discourse to get their ideas accepted and at the same time how this works to construct knowledge and sustain and change disciplinary communities. This is a key issue of academic discourse analysis and of the study of human interaction more generally, as Stubbs (1996: 21) observes: The major intellectual puzzle in the social sciences is the relation between the micro and the macro. How is it that routine everyday behavior, from moment to moment, can create and maintain social institutions over long periods of time? In this enterprise discourse analysis, particularly text-based forms of genre analysis, has become established as the most widely used and productive methodology.

How Is Academic Discourse Studied? Discourse analysis comprises a broad collection of methods for studying language in action, looking at texts in relation to the social contexts in which they are used. Because language is an irreducible part of social life, this broad definition has been interpreted in various ways across the social sciences. In academic contexts, however, it has tended to be a methodology which focuses on concrete texts rather than institutional social practices. In particular, it has largely taken the form of focusing on particular academic genres such as the research article, conference presentation and student essay. Genre analysis can be seen as a more specific form of discourse analysis which focuses on any element of recurrent language use, including grammar and lexis, is relevant to the analyst’s interests. As a result, genre analysis sees texts as representative of wider rhetorical practices and so has the potential to offer descriptions and explanations of both texts and the communities that use them. Genres are the recurrent uses of more-or-less conventionalized forms through which individuals develop relationships, establish communities and get things done using language. Genres can thus be seen as a kind of tacit contract between writers and readers, which influence the behaviour of text 174

HylandK_Chap11_Finals.indd 174

12/1/2010 8:05:30 AM

Academic Discourse

producers and the expectations of receivers. By focusing on mapping typicality, genre analysis thus seeks to show what is usual in collections of texts and so helps to reveal underlying ideologies and Discourses (Gee 1999) and the preferences of disciplinary communities. These approaches are influenced by Halliday’s (1994) view of language as a system of choices which link texts to particular contexts through patterns of lexico-grammatical and rhetorical features and by Swales’ (1990) observation that these recurrent choices are closely related to the work of particular discourse communities whose members share broad social purposes. One of the most productive applications of discourse analysis to academic texts has been to explore the lexico-grammatical regularities of particular genres to identify their structural identity. Analysing this kind of patterning has yielded useful information about the ways texts are constructed and how we recognize coherent patterning of text elements. Some of this research has followed the move analysis work pioneered by Swales (1990), which seeks to identify the stages of particular institutional genres and the constraints on typical move sequences. Moves are the rhetorical steps which writers or speakers routinely use to develop their social purposes, and recent work on academic genres has produced descriptions of dissertation acknowledgments (Hyland 2004a), the methods sections in research articles (Bruce 2009) and the peer seminar (Aguia 2004). While analysing schematic structures has proved a useful way of looking at texts, analysts are increasingly aware of the dangers of oversimplifying by assuming blocks of texts to be mono-functional and ignoring writers’ complex purposes and ‘private intentions’ (Bhatia 1999). There is also the problem of validating analyses to ensure they are not simply products of the analyst’s intuitions (Crooke 1986). Transitions from one move to another are always motivated outside the text as writers respond to their social context, but analysts have not always been able to identify the ways these shifts are explicitly signalled by lexico-grammatical patterning (e.g. Paltridge 1994). Consequently, attention has turned to particular features of specific genres, either grammatical, such as circumstance adverbials in student presentations (Zareva 2009), functional, like hedging in research articles (Hyland 1998), or rhetorical, such as evaluation in book reviews (Hyland and Diani 2009). Corpora are increasingly used to identify frequent choices in different modes and genres, with recent studies exploring the common four-word collocations, or lexical bundles, which are typical in undergraduate textbooks (Biber 2006) and student dissertations (Hyland 2008). While text analysis is an essential part of discourse analysis, discourse analysis is not merely the linguistic analysis of texts. In academic contexts the interpretive and qualitative study of both texts and users has begun to grow in recent years to establish the ways that texts are firmly embedded in the cultures 175

HylandK_Chap11_Finals.indd 175

12/1/2010 8:05:31 AM

Bloomsbury Companion to Discourse Analysis

and activities in which their users participate. One example is Prior’s (1998) study of the contexts and processes of graduate student writing at a US university. Drawing on transcripts of seminar discussions, student texts, observations of institutional contexts, tutor feedback and interviews with students and tutors, Prior provides an in-depth account of the ways students in four fields negotiated their writing tasks and so became socialized into their disciplinary communities. In another study, Li (2006) shows how advice from supervisors, a journal editor and reviewers helped guide a Chinese doctoral student of physics through six drafts and several resubmissions before her paper was finally accepted for publication. Ethnographic-oriented studies have also explored the literate cultures of academics themselves. Perhaps the best known of these is Swales’ (1998) ‘textography’ of his building at the University of Michigan. Swales makes greater use of analyses of texts and systems of texts in his approach than many ethnographies, combining discourse analyses with extensive observations and interviews. Together these methods provide a richly detailed picture of the professional lives, commitments and projects of individuals in three diverse academic cultures working in the building: the computer centre, the Herbarium and the university English Language Centre. The interplay of different types of data allows us to see how the multiple influences of academic practices, peers, mentors and personal experiences all contribute to their texts and experiences as academic writers (Paltridge 2008). Finally, studies conducted from a Critical perspective have focused on how social relations, identity, knowledge and power are constructed through written and spoken texts in disciplines, schools and classrooms. Distinguished by an overtly political agenda, Critical Discourse Analysis (CDA) has attempted to show that the discourses of the academy are not transparent or impartial means for describing the world but work to construct, regulate and control knowledge, social relations and institutions. Particular literacy practices possess authority because they represent the currently dominant ideological ways of depicting relationships and realities and these authorized ways of seeing the world exercise control of academics and students alike. Lillis (2001), for instance, shows how this can create tensions for students in coping with university literacy demands, while Flowerdew’s (1999) research suggests similar concerns among non-native English scholars.

What Do We Know about Academic Discourse? Together these different approaches aim at capturing thicker descriptions of language use in the academy, producing a rich vein of research findings which continues to inform both teaching and our understanding of the practices of 176

HylandK_Chap11_Finals.indd 176

12/1/2010 8:05:31 AM

Academic Discourse

disciplinary knowledge-making. The scale of this research makes it difficult to summarize, but we can identify four main findings: (1) That academic genres are persuasive and systematically structured to secure readers’ agreement; (2) That these ways of producing agreement represent disciplinary specific rhetorical preferences; (3) That language groups have different ways of expressing ideas and structuring arguments; (4) That academic persuasion involves interpersonal negotiations as much as convincing ideas.

Academic Texts Are Structured for Persuasive Effect All academic texts are designed to persuade readers of something: of the knowledge claim at the heart of a research article or dissertation; of an evaluation of others’ work in a book review, or of one’s understanding and intellectual autonomy in an undergraduate essay. To accomplish these various purposes, writers tend to draw on the same repertoire of linguistic resources for each genre again and again. This is, in part, because writing is a practice based on expectations. The process of writing involves creating a text that the writer assumes the reader will recognize and expect and the process of reading involves drawing on assumptions about what the writer is trying to do. Hoey (2001) says that this is like dancers following each other’s steps, each assembling sense from a text by anticipating what the other is likely to do by making connections to prior texts. While writing, like dancing, allows for creativity and the unexpected, established patterns form the basis of any variations. This schema of prior knowledge, acquired through formal learning and repeated experiences with texts, allows writers and speakers to express themselves appropriately and effectively, drawing on conventions for organizing messages so that their readers can recognize their purpose and follow their ideas. The research article, for instance, is a genre which restructures the processes of thought and the research it describes to establish a discourse for scientific fact-creation. Language becomes a form of technology in this highly refined genre as it attempts to present interpretations and position participants in particular ways as a means of establishing knowledge. A range of spoken and written academic genres have been studied in recent years. These include student dissertations (Bunton 2002), conference presentations (Carter-Thomas and Rowley-Jolivet 2001) and grant proposals (Connor and Upton 2004). This research demonstrates the distinctive differences in the genres of the academy where particular purposes and audiences lead writers 177

HylandK_Chap11_Finals.indd 177

12/1/2010 8:05:31 AM

Bloomsbury Companion to Discourse Analysis

Table 11.1 Selected features in research articles and textbooks per 1,000 words Research Articles University Textbooks

Hedges 15.1 8.1

Self-mention

Citation

Transitions

3.9 1.6

6.9 1.7

12.8 24.9

to employ very different rhetorical choices. Table 11.1, for example, compares frequencies for different features in a corpus of 240 research articles and 56 textbooks. We can see considerable variation in these features across the two genres. The greater use of hedging underlines the need for caution and opening up arguments in the research papers compared with the authorized certainties of the textbook, while the removal of citation in textbooks shows how statements are presented as facts rather than claims grounded in the literature. The greater use of self-mention in articles points to the personal stake that writers invest in their arguments and their desire to gain credit for claims. The higher frequency of transitions, which are conjunctions and other linking signals, in the textbooks is a result of the fact that writers need to make connections far more explicit for readers with less topic knowledge.

Academic Texts Represent Discipline-specific Modes of Argument A second finding of research is that successful academic writing depends on the individual writer’s control of the epistemic conventions of a discipline, what counts as appropriate evidence and argument, and that this differs across fields. Research on language variation across the disciplines is now one of the more fruitful lines of research and one of the dominant paradigms in EAP (e.g. Hyland 2004b; Hyland and Bondi 2006). The idea of discipline is rather nebulous (Hyland 2009; Mauranen 2006), but captures how individuals use and respond to language as members of social communities. Challenged by postmodernism, interdisciplinary research and the emergence of modular degrees, the notion of discipline is often questioned (e.g. Gergen and Thatchenkery 1996). But while boundaries are never stable nor objects of study immutable, discipline is a notion with remarkable persistence. The distinctiveness of disciplines, however, can be informed by study of rhetorical practices. This is because successful academic writing depends on writers’ projections of a shared professional context as they seek to embed their writing in a particular social world which they reflect and conjure up through approved discourses. 178

HylandK_Chap11_Finals.indd 178

12/1/2010 8:05:31 AM

Academic Discourse

Essentially, we can see disciplines as language-using communities and this helps us join writers, texts and readers together. Communities provide the context within which we learn to communicate and to interpret each other’s talk, gradually acquiring the specialized discourse competencies to participate as members. So we can see disciplines as particular ways of doing things – and predominantly as particular ways of using language to engage with others. Speakers and writers thus make language choices to gain support, express collegiality and resolve difficulties in ways which fit the community’s assumptions, methods, and knowledge. Wells (1992) puts this succinctly: Each subject discipline constitutes a way of making sense of human experience that has evolved over generations and each is dependent on its own particular practices: its instrumental procedures, its criteria for judging relevance and validity, and its conventions of acceptable forms of argument. In a word each has developed its own modes of discourse. So, disciplines structure research within wider frameworks of beliefs and provide the conventions and expectations that make texts meaningful. In the sciences new knowledge is accepted through experimental proof. Science writing reinforces this by highlighting a gap in knowledge, presenting a hypothesis related to this gap, and then reporting experimental findings to support this. The humanities, on the other hand, rely on case studies and narratives while claims are accepted on strength of argument. The social sciences fall between these poles because in applying scientific methods to less predictable human data they have to give more attention to explicit interpretation. In other words, academic discourse helps to give identity to a discipline and analyses of texts help reveal the distinctive ways disciplines have of asking questions, addressing a literature, criticizing ideas and presenting arguments. Research has discovered considerable rhetorical variation across a range of features in genres such as scientific letters (Hyland 2004b), writing assignments (Gimenez 2009) and PhD dissertations (Hyland 2004c). One of the most striking differences in how language differs across fields is the use of hedges such as possible, might, probably and so on. These function to withhold complete commitment to a proposition, implying that a claim is based on plausible reasoning rather than certain knowledge. They indicate the degree of confidence the writer thinks it might be wise to give a claim while opening a discursive space for readers to dispute interpretations (Hyland 1998). Because they represent the writer’s direct involvement in a text, something that scientists generally try to avoid, they are twice as common in humanities and social science papers than in hard sciences. One reason for this is there is less control of variables, more diversity of research outcomes and fewer clear bases for accepting claims than in the sciences. Writers 179

HylandK_Chap11_Finals.indd 179

12/1/2010 8:05:31 AM

Bloomsbury Companion to Discourse Analysis

cannot report research with the same confidence of shared assumptions so papers rely far more on recognizing alternative voices. Arguments have to be expressed more cautiously by using more hedges. In the hard sciences positivist epistemologies mean that the authority of the individual is subordinated to the authority of the text and facts are meant to ‘speak for themselves’. Writers, therefore, often disguise their interpretative activities behind linguistic objectivity. They downplay their personal role to suggest that results would be the same whoever conducted the research. The less frequent use of hedges is one way of accomplishing this.

Different Cultures Have Different Language Schemata Academic discourse analysis has also pointed to cultural specificity in rhetorical preferences (e.g. Connor 2002). Although a controversial term, one version of culture regards it as an historically transmitted and systematic network of meanings, which allow us to understand, develop and communicate our knowledge and beliefs about the world. Culture is seen as inextricably bound up with language (Kramsch 1993), so that cultural factors have the potential to influence perception, language, learning and communication. Although it is far from conclusive, discourse analytic research suggests that the schemata of L2 and L1 writers differ in their preferred ways of organizing ideas which can influence academic writing (e.g. Hinkel 2002). These conclusions have been supported by a range of studies into different genres over the past decade (e.g. Duszak 1997; Yakhontova 2002). Much of this work has focused on student genres and has identified a range of different features in first and second language writing in English, particularly, the ways writers incorporate material into their writing, how they orientate to readers through attention-getting devices and estimates of reader knowledge, and differences in the use of overt linguistic features (such as less subordination, more conjunction, less passivization, fewer free modifiers, less noun-modification, less specific words, less lexical variety, predictable variation and a simpler style) (e.g. Grabe and Kaplan 1996: 239). Critics point out, however, that because contrastive rhetoric starts from an assumption of difference, it has ‘tended to look at L2 writing ... mainly as a problem of negative transfer of L1 rhetorical patterns to L2 writing’ (Casanave 2004: 41). This not only sees L2 writing as a deficit, but runs the risk of ignoring the rich and complex histories of such students’ literacies and what they bring to the L2 classroom (e.g. Horner and Trimbur 2002). Equally seriously, much of the contrastive rhetoric research into discourse assumes a ‘received view of culture’ which unproblematically identifies cultures with national entities and emphasizes predictable consensuality within 180

HylandK_Chap11_Finals.indd 180

12/1/2010 8:05:32 AM

Academic Discourse

cultures and differences across them (e.g. Atkinson 2004). However, it is fair to say that, compared with many languages, academic writing in English tends to: z be more explicit about its structure and purposes with constant preview-

ing and reviewing; z employ more, and more recent, citations; z be less tolerant of digressions; z be more cautious in making claims, with considerable use of mitigation

and hedging; z use more sentence connectors to show explicitly how parts of the text link

together. While we cannot simply predict the ways people are likely to write on the basis of assumed cultural traits, discourse studies have shown that students’ first language and prior learning come to influence ways of organizing ideas and structuring arguments when writing in English at university.

Academic Argument Involves Interpersonal Negotiations Much recent work has focused on how persuasion in various genres is not only accomplished through the ways ideas are presented, but also by the construction of an appropriate authorial self and the negotiation of participant relationships. While once considered a self-evidently objective and impersonal form of discourse, academic writing is now widely considered to be a persuasive endeavour. Academics do not simply produce texts that plausibly represent an external reality, but use language to acknowledge, construct and negotiate social relations. Discourse analysis has helped to show how writers offer a credible representation of themselves and their work by claiming solidarity with readers, evaluating their material and acknowledging alternative views. Interaction in academic writing essentially involves ‘positioning’, or adopting a point of view in relation to both the issues discussed in the text and to others who hold points of view on those issues. In persuading readers of their claims writers must display a competence as disciplinary insiders that is, at least in part, achieved through a writer–reader dialogue which situates both their research and themselves. As this view gains greater currency, considerable attention has turned to the features which help realize this interpersonal and evaluative dimension of academic texts. Genres such as undergraduate lectures (Morell 2004), conference monologues (Webber 2005) and book reviews (Hyland and Diani 2009) have been explored from this perspective. Several frameworks have 181

HylandK_Chap11_Finals.indd 181

12/1/2010 8:05:32 AM

Bloomsbury Companion to Discourse Analysis

been proposed to analyse the linguistic resources employed in this way, with research conducted under labels such as ‘evaluation’ (Hunston and Thompson 2000), ‘appraisal’ (Martin and White 2005), ‘metadiscourse’ (Hyland 2005a) and ‘stance and engagement’ (Hyland 2005b). Stance and engagement refers to various rhetorical features which help construct both writers and readers. Stance is an attitudinal dimension which includes features which refer to the ways writers present themselves and convey their judgements, opinions and commitments, either intruding to stamp their personal authority onto their arguments or stepping back to disguise their involvement. Engagement, in contrast, is an alignment dimension where writers acknowledge and connect to others, recognizing the presence of their readers by focusing their attention, acknowledging their uncertainties and including them as discourse participants. Metadiscourse, on the other hand, seeks to offer a more comprehensive way of examining interaction in academic argument, broadening the scope of interactional resources to also include features such as conjunctions, framing devices and glosses on content. While these are often considered as simply helping to tie texts together, they have an important role in relating a text to a community.

A Sample Study: Citations The importance of interaction in academic writing and the variation of differences in disciplinary uses of language are evident in citation practices and illustrated in a study I conducted of 80 research articles in 8 disciplines comprising 700,000 words and interviews with specialist informants (Hyland 2004). The inclusion of references to the work of other authors is obviously central to academic persuasion. Not only does it help writers to establish a persuasive context and social framework for their arguments by showing how a text depends on the understandings and previous work in a discipline, but it also displays the writer’s status as an insider. It helps align him or her with a particular community or orientation and establish a credible writer ethos, confirming that this is someone who is aware of, and is knowledgeable about, the topics, approaches and issues which currently inform the field. This helps to explain why I found 70 citations per paper, but because discourse communities see the world in different ways they also write about it in different ways. Table 11.2 shows that two-thirds of all the citations in the corpus were in the philosophy, sociology, marketing and applied linguistics papers, twice as many as in the science disciplines. These differences reflect the extent writers can assume a shared context with readers. In Kuhn’s ‘normal science’ model, natural scientists produce public knowledge through relatively steady cumulative growth. Results throw 182

HylandK_Chap11_Finals.indd 182

12/1/2010 8:05:32 AM

Academic Discourse

Table 11.2 Rank order of citations by discipline Discipline Sociology Philosophy Applied Linguistics Marketing

per 1,000 words 12.5 10.8 10.8 10.1

Discipline Biology Electronic Engineering Mechanical Engineering Physics

per 1,000 words 15.5 8.4 7.3 7.4

up more questions to be answered by further research so writers do not need to report research with extensive referencing. Those who read the papers are often working on the same problems and are familiar with the earlier work and with the methodologies used. In the humanities and social sciences, on the other hand, the literature is more dispersed and the readership more heterogeneous, so writers cannot presuppose a shared context but have to build one far more through citation. Biology spoils the rather neat division between hard and soft knowledge practices by having the highest citation count per 1,000 words. Interestingly, this is largely due to a high proportion of self-citation with 13 per cent of all citations to the current author compared with about 6 per cent in the other disciplines. This is a recognition of ownership of ideas which amounts to knowledge as private property in biology unknown in other sciences (Hyland 2004b). There are also major differences in the ways writers report others’ work as writers in different fields draw on very different sets of reporting verbs to refer to their literature (Hyland 2004b). Among the higher frequency verbs, almost all instances of say and 80 per cent of think occurred in philosophy and 70 per cent of use in electronics. It turns out, in fact, that engineers show, philosophers argue, biologists find and linguists suggest. These preferences reflect broad disciplinary purposes. So, the soft fields largely use verbs which refer to writing activities, like discuss, hypothesize, suggest, argue. These involve the expression of arguments and allow writers to discursively explore issues while carrying a more evaluative element in reporting others’ work. Engineers and scientists, in contrast, prefer verbs which point to the research itself like observe, discover, show, analyse and calculate, which represent real world actions. This helps scientists represent knowledge as proceeding from impersonal lab activities rather than from the interpretations of researchers. These disciplinary differences suggest that citations are related to community-specific norms of effective argument which involve appropriate interpersonal interactions. Citation practices, like many other features, show that writers frame their studies for colleagues in ways that represent inquiry patterns and conventions of argument which reinforce ideologies of objectivity or engagement. 183

HylandK_Chap11_Finals.indd 183

12/1/2010 8:05:32 AM

Bloomsbury Companion to Discourse Analysis

Conclusion Discourse studies have provided analysts with insights into the ways academics and students actively engage in knowledge construction as members of professional groups, revealing something of how their discoursal decisions are socially grounded in the knowledge structures and rhetorical repertoires of their disciplines. As this research continues to grow, we can anticipate an ever increasing broadening of studies beyond texts to the talk and contexts which surrounds their production and use, beyond the verbal to the visual, and beyond tertiary to school and professional contexts.

Key Readings Biber, D. (2006), University Language: A Corpus-based Study of Spoken and Written Registers. Amsterdam: Benjamins. Hyland, K. (2004), Disciplinary Discourses. Ann Arbor, MI: University of Michigan Press. Hyland, K. (2009), Academic Discourse. London: Continuum. Swales, J. (1998), Other Floors, Other Voices: A Textography of a Small University Building. Mahwah, NJ: Erlbaum. Swales, J. (2004), Research Genres. New York: Cambridge University Press.

184

HylandK_Chap11_Finals.indd 184

12/1/2010 8:05:32 AM

12

Discourse and the Workplace Janet Holmes

Chapter Overview Introduction Current Research on Spoken Workplace Discourse A Sample Study New Directions in Workplace Discourse Research Transcription Key Notes Key Readings

185 186 192 197 198 198 198

Introduction Research on workplace discourse has increased considerably in the last ten years. This chapter is necessarily selective, though hopefully it indicates something of the breadth, as well as the fascination, of current research on workplace discourse.1 Early research on workplace discourse focused on doctor–patient interaction and legal language, especially courtroom discourse, and these remain areas of interest for many researchers (see Olsson this volume; Halkowski this volume). But the scope of research has broadened considerably in the past 30 years and now encompasses many different types of institutional and non-institutional contexts, as well as many different aspects of workplace interaction, such as the structure of talk in negotiation, the role of humour and small talk at work, the construction of professional identities in the workplace and the place of email in workplace interaction. While recognizing that much valuable work has been undertaken on written discourse in workplace contexts (e.g. Gunnarsson 2009), this chapter focuses on talk at work. The next section provides an overview and discussion of current 185

HylandK_Chap12_Finals.indd 185

12/1/2010 8:05:52 AM

Bloomsbury Companion to Discourse Analysis

research on spoken workplace discourse. Overall, this discussion focuses on qualitative rather than quantitative approaches, as illustrated in more detail in the third section which uses a brief sample study drawn from the work of the Wellington Language in the Workplace Project (LWP) to illustrate some of the points discussed. The chapter ends with some brief reflections on potential future directions for the analysis of workplace discourse.

Current Research on Spoken Workplace Discourse Following a brief note on methodology, this section discusses research on workplace discourse in three broad categories: first, different types of workplace interaction, second, two well-established, broad sociolinguistic dimensions of analysis – power and solidarity; and third, two areas of social variation, gender and ethnicity in the workplace.

Methodology: A Brief Note While ethnographic approaches involving participant observation and interviews often provide supplementary data, the prevailing data collection methodology in workplace discourse research involves recording naturally occurring talk in ‘authentic’ situations. Early researchers, using a predominantly Conversation Analysis (CA) approach to analysis, audio-recorded relatively formal interactions such as job interviews, interactions between health visitors and clients, and service encounters. Others focused on collecting audio- and sometimes video-recordings of workplace meetings (Boden 1994; e.g. BargielaChiappini and Harris 1996). Gradually, the scope of recording expanded, and Clyne (1994) describes a project where factory workers from diverse ethnolinguistic backgrounds carried microphones to record their everyday workplace talk. This methodology has since been expanded and adapted for a wide range of different workplaces (Holmes and Stubbe 2003), and it is currently the predominant approach used in collecting workplace interaction (Koester 2006; Mullany 2007; Richards 2006; Schnurr 2009). In terms of analysis, current research exemplifies a range of approaches to analysing workplace discourse, ranging from ethnographic approaches, through micro-level description of the details of talk provided by CA approaches, to the politically motivated framework adopted by Critical Discourse Analysts (CDA), and the more quantitative approach of corpus analysts. These are discussed in greater detail elsewhere in this volume.2 This chapter concentrates instead on giving readers a sense of the range of contexts and the dimensions of analysis that researchers in workplace discourse have found useful. One distinction 186

HylandK_Chap12_Finals.indd 186

12/1/2010 8:05:53 AM

Discourse and the Workplace

which has proved valuable is the distinction between the transactional (or referential) dimension and the relational (or interpersonal or affective) dimension of meaning in talk. Analysts emphasize the fact that the distinction is simply a useful heuristic tool, since every interaction has elements of both dimensions of meaning as will be apparent in the excerpts discussed below. In current research, the distinction facilitates the analysis of the complexities of a wide range of different kinds of workplace interaction, such as small and large meeting discourse and talk in service encounters, as well as the construction of professional identity, including leadership and expert identity. Some of this complexity will emerge in the discussion which follows.

Different Types of Workplace Interaction In this section, research on workplace interaction in different kinds of settings is discussed. The main focus is on meetings, probably the most extensive area of workplace discourse analysis, followed by a brief discussion of research on service encounters and other occupational genres. White collar professionals spend a very large proportion of their time in meetings. Different studies (involving different methods of calculation) indicate that meetings occupy anything between 25 per cent and 80 per cent of the work time of the white-collar workforce. It is not surprising, then, that discourse analysts have devoted a good deal of attention to meeting talk, examining the discursive strategies used in the management of meetings (e.g. Bargiela-Chiappini and Harris 1996), and the complexities of how things get accomplished interactionally through meeting talk (e.g. Drew and Heritage 1992; Firth 1995; Geyer 2008; Sarangi and Roberts 1999). More specifically, those using a CA approach have identified patterns which help define a meeting, such as the structure of the opening and closing phases and the distribution of turns (Boden 1994; Mirivel and Tracy 2005), while others have discussed discursive ways in which topics are demarcated in meetings (Bargiela-Chiappini and Harris 1996). Research by our Wellington LWP team notes that conventions for formal meeting openings and closings often differ for different cultural groups, and such differences may influence workplace meeting norms, as was evident in the Māori workplaces researched (Holmes, Marra and Schnurr 2008). Researchers analysing the discourse patterns in meetings between Chinese and British business people (Spencer-Oatey and Xing 2003), between Chinese and Englishspeaking Westerners (Bilbow 1998) and between Japanese and American business people (Yamada 1997) have similarly identified culturally significant contrasting patterns in the appropriate ways of interacting in such meetings. For example, Chinese business meetings generally maintain a relatively high level of formality and the opportunity to deliver a formal speech stating 187

HylandK_Chap12_Finals.indd 187

12/1/2010 8:05:53 AM

Bloomsbury Companion to Discourse Analysis

one’s position is considered very important and a sign of respect. By contrast, Western business meetings often favour informality; indeed ‘dispensing with formalities’ is regarded positively as a sign of good rapport. Such differences can result in offence in intercultural meetings, and they led to a significant communication breakdown in the case study documented by Spencer-Oatey and Xing (2003). Workplace interactions between American and Japanese workers in car-manufacturing factories have been studied by Sunaoshi (2005), with a focus on how workers from different cultural backgrounds negotiate through to understanding. And there is considerable work by interactional sociolinguists on intercultural communication, focusing especially on the significance for migrants of miscommunication in crucial work sites such as job and promotion interviews (e.g. Campbell and Roberts 2007). (See also Corbett this volume.) The importance of the relational dimension in meeting talk has been given quite explicit attention by some discourse analysts (Holmes and Stubbe 2003; Koester 2006; Mullany 2007). It is common for meeting openings to be preceded by solidarity-building social talk, for example, as people wait for the meeting participants to assemble (Mirivel and Tracy 2005). But more subtly, relational aspects of interaction are also pervasively evident throughout meeting talk on predominantly transactional issues, as Geyer (2008) illustrates in her detailed analysis of facework in six teacher meetings in Japanese secondary schools. Adopting a dynamic postmodern approach, she conceptualizes ‘face’ as a ‘speaker’s interactional social image’ (2008: 6), and examines how ‘an interlocutor ascribes and is ascribed multiple discursive and social identities which in turn can invoke multiple faces’ (2008: 6–7). Meetings are also sites for constructing and enacting facets of professional identity. While there is some research on aspects of professional identity in meetings between just two or three participants (Koester 2006; Vine 2004), most analyses focus on relatively large group meetings. In such contexts, the role of the meeting chair provides many opportunities for asserting status and ‘doing power’, since the chair has the right to declare the meeting open, to move discussion to new agenda items, to summarize progress, to ratify decisions and to close the meeting (e.g. Bilbow 1998; Holmes and Stubbe 2003). Discourse analysts have also examined features of service encounters involving different interactional contexts, ranging from corner shops, through bookshops and hair salons, to supermarkets, and including telephone transactions. Research has focused on the structure and dynamics of service transactions, including the role of small talk, humour and other politeness devices, and the discourse analysis sometimes involves consideration of the skilful use of more than one language to accomplish interactional goals in service encounters (e.g. Kuiper 2009). The ways in which discourse is used to construct and negotiate different aspects of social identity by participants has also been analysed in this context. In Galicia, for instance, Prego-Vazquez (2007) describes 188

HylandK_Chap12_Finals.indd 188

12/1/2010 8:05:54 AM

Discourse and the Workplace

how customers switch between Spanish and colloquial and dialectal Galician in order to negotiate professional and personal identities in interaction with borough officials concerned with water and sewage treatment services. Finally, there is a small body of research which focuses on more specialized occupational genres, such as the discourse of auctioneers, sportscasters and weather forecasters, and the structural features of airline pilot talk (Nevile 2007) including how these contribute to the construction of complex occupational identities (Ashcraft 2007).

Power and Solidarity in Workplace Interaction These two dimensions of analysis provide a framework for discussing research on how people perform leadership at work, as well as for considering the role of humour and narrative in workplace interaction. There is a huge literature on leadership, but the ‘turn to discourse’ in this area has been relatively recent, emerging from social constructionist approaches. Earlier research used questionnaire and interview data rather than examining how leaders actually talk at work. Since, according to one study, leaders spend on average ‘between 62% and 89% of their time in face-to-face communication’, and ‘between one-third and two-thirds of their time communicating with their subordinates’ in the workplace (Gardner, Raschka and Sercombe 1996: 153), a focus on their workplace talk seems well justified. The research of the LWP team has been seminal in this area (see publications on webpage: www.victoria.ac.nz/lals/lwp). Our detailed discourse analyses support claims by leadership scholars that effective leadership talk demonstrates a range of diverse competencies, including transactional skills oriented to achieving workplace objectives, and relational skills which take account of interpersonal aspects of communication, such as establishing rapport with colleagues, and attenuating directives. Social constructionist approaches thus emphasize the complexities and ambiguities associated with accomplishing leadership, as well as the importance of discourse in managing meaning. Current theories of leadership highlight the importance both of assertiveness and authority, qualities normatively associated with masculine styles of interaction, as well as relational skills, attributes normatively associated with more feminine interactional styles. Thus both women and men must negotiate a complex path through the social expectations which surround the leadership role to construct a satisfactory identity in their specific communities of practice (Baxter 2010; Holmes 2006; Mullany 2007; Schnurr 2009). In this process, effective leaders draw from a wide and varied discursive repertoire, selecting appropriate discursive strategies in response to particular interactional contexts. Using a feminist poststructuralist theoretical perspective to analyse her data, Baxter 189

HylandK_Chap12_Finals.indd 189

12/1/2010 8:05:54 AM

Bloomsbury Companion to Discourse Analysis

(2010) argues that effective leaders need to juggle between different and competing subject positions in order to fulfil a range of business goals, and that female leaders have added demands in order to sustain a credible set of professional identities. Gender in workplace discourse research is discussed further below. Research on speech functions such as directives also involves consideration of the power dimension in workplace discourse analysis. Good management entails getting people to do things at work. Vine (2004) examines the range of ways in which directives are expressed in the interactions of four senior employees in a government department. She demonstrates, in particular, that the force of any directive, and especially its interpersonal or relational impact, is greatly influenced by its precise discursive positioning in a particular social context, with the relevant context sometimes extending over several meetings. Similar conclusions have resulted from the analysis of workplace disagreement and negotiation (e.g. Geyer 2008; Koester 2006; Richards 2006). A related strand of speech function research has focused on the challenges for those from other cultures in appropriately expressing problematic or challenging speech acts such as refusals and complaints, especially when these are directed upwards to those in positions of greater power (e.g. Clyne 1994). Turning to the solidarity dimension, humour has attracted a good deal of attention from workplace discourse analysts (see Schnurr 2009 for a recent review). Early research in this area often argued that workplace humour benefited employment relationships by increasing job satisfaction, creativity and even productivity. In contexts as diverse as hospitals paramedical departments, hotel kitchens and police departments, humour was shown to have beneficial effects. Looking at the ‘darker side’ of workplace humour, others argued that it could be used manipulatively, as a control mechanism, for example, by the chair in white-collar business meetings (Mullany 2007), or by the management in a factory to encourage conformity to group norms. Alternatively, humour was construed as a strategy for expressing resistance. In a white-collar, commercial context, Rodrigues and Collinson (1995) demonstrated that Telecom employees in Brazil not only used humour (and particularly cartoons) as a safety valve for channelling emotions and expressing dissatisfaction, but also as a weapon of contestation and a means to effect change. More recently, humour has been considered as a component of workplace culture (e.g. Holmes and Stubbe 2003; Schnurr 2009). To a greater or lesser extent, every organization develops a distinct workplace culture, and particular workplace teams often develop as distinctive communities of practice, with particular ways of doing things and systems of shared understandings within an organization. Humour frequently plays a part in this process. IT companies, for instance, are often characterized by a distinctively masculine, contestive and challenging style of interactive humour. In British academic 190

HylandK_Chap12_Finals.indd 190

12/1/2010 8:05:54 AM

Discourse and the Workplace

and research team contexts, (2006) illustrates a more affiliative and collaborative style of humour in meetings, contrasting with different styles in different communities of practice in less formal contexts. Others have examined how workplace leaders and their team members often use humour to collaborate in constructing not only a particular type of workplace culture, but also a leadership style appropriate to their particular community of practice (e.g. Schnurr 2009). Research on workplace narratives has similarly examined their contribution to the construction of complex professional identities (e.g. Mullany 2006). While most researchers have focused on the construction of workplace identities through individual’s narratives, some have shown how groups (often jointly) use stories to construct themselves as a productive and professional team, or a competitive squad, as a ‘family business’ or a streamlined organization (e.g. Richards 2006). Conversely, like humour, narrative provides a subtle means of contesting or subverting the prevailing organizational ethos or workplace culture, through stories which present an alternative reality. And workplace stories also provide a means of constructing the professional identities of others as effective or inadequate, competent or incompetent team members. Narrative can also make an important contribution to gendered discourse, reinforcing stereotypical constructs of masculinity, such as hierarchy, structure, dominance, competitiveness, aggression and goal-oriented action, or normatively feminine dimensions such as egalitarianism, collaboration and cooperation (Baxter 2010; Holmes 2006). An interesting foray into a blue collar work context, for instance, demonstrated how three white, working class, male builders constructed their professional identities through collaborative narratives while travelling in a truck between different building sites (Baxter and Wallace 2009). The next section focuses explicitly on gender in workplace discourse.

Gender and Ethnicity in the Workplace Inevitably gender has already crept into the discussion of workplace discourse analysis at several points, reflecting the steady increase in research in this area. The ways in which discourse contributes to the construction of gendered identities at work has attracted particular attention. Again the predominant paradigm is a dynamic, social constructionist approach, recognizing that both women and men operate within the constraints of overarching society-wide behavioural norms and expectations in constructing their gender identity in the workplace. The challenges which people often face in enacting workplace roles involving conflicting sets of norms for speaking have been identified by a number of language and gender researchers. The pervasive power of gender 191

HylandK_Chap12_Finals.indd 191

12/1/2010 8:05:54 AM

Bloomsbury Companion to Discourse Analysis

norms is particularly evident when the spotlight is turned on those contesting them in some way. So, for example, the policewomen observed and interviewed by McElhinney (1995) commented that they deliberately repressed any emotional (normatively feminine) response when dealing with stressed citizens, and assumed a demeanour they considered appropriate to their professional role. There is also research which indicates the extent to which discourse contributes to the construction of gendered workplaces. Many occupations such as IT and engineering, and their corresponding workplaces are still male-dominated, while areas such as education and nursing tend to employ more women, with consequences for the discursive norms constructed in these areas. Focusing on a non-Western cultural and linguistic context, Philips (2007) describes the ways in which a Tongan women’s work group producing decorated bark cloth, a ‘quintessentially female work activity’ (2007: 67), engage in discourse activities which richly enact diverse aspects of Tongan gender ideology. The women’s place in Tongan society as well as within a global economy is thus constructed and reinforced both through their work and their talk. Few studies have explicitly focused on ethnicity as a component in workplace discourse and this is an area which may be expected to develop in future research. Approaching ethnicity within a constructionist framework involves regarding it as a dynamic and active process enacted in ongoing interaction and ethnic boundaries as negotiated by individuals and groups in response to their evolving social roles and circumstances. While researchers in non-Western societies have undertaken valuable research within such a framework, especially on the discourse of women at work (e.g. Philips 2007), there is little on the relevance of ethnicity in Western work contexts. The LWP team has begun work in this area in the New Zealand context using the concept of the ethnicized workplace or community of practice (Holmes et al. 2008; Holmes, Marra and Vine forthcoming), a place where ethnicity acts as a taken-for-granted backdrop, crucial for interpreting everyday communication; ethnic values underpin the guidelines or norms which influence the way people interact, and the ways in which they construct different aspects of their identity, including their ethnicity. Ethnicity permeates workplace communication in such organizations. This is illustrated in the sample study in the next section which brings together a number of the concepts explored in the preceding sections.

A Sample Study This section provides a brief case study of Quentin, a senior manager in an ‘ethnicised community of practice’, a Māori workplace with transactional goals 192

HylandK_Chap12_Finals.indd 192

12/1/2010 8:05:54 AM

Discourse and the Workplace

related to producing good quality commercial outputs, while also explicitly committed to furthering Māori objectives and promoting Māori values. The study illustrates many of the concepts and dimensions of analysis introduced in the previous discussion including meeting discourse, transactional and relational dimensions of analysis, power and solidarity, and the relevance of gender and ethnicity in the analysis of workplace discourse (see also Holmes et al. forthcoming). Quentin discursively enacts leadership by consistently providing direction to his team, and identifying clear goals and objectives. But perhaps his most valuable contribution to the organization lies in his ability to work through the detailed practical steps required to achieve the organization’s transactional objectives. This practical orientation is evident in his contributions to all meetings, both large and small. Example 1 provides an instance of Quentin’s attention to what can be done to facilitate and speed up progress on a particular project. Example 1 Context: meeting between Quentin and one of his senior team members 1 Quentin: //I’m just trying to think of ways eh yeah yeah\ 2 and and and getting value out of the time 3 of whether we bring some [professionals] in 4 and we have a discussion that we actually record 5 Paula: /oh yeah yeah yeah ways of making it fast 6 having a discussion\\ ... 7 Quentin: you know give them that 8 [laughs]: I just see that this yeah: yeah yeah 9 I’m just thinking of ways in which it ight help and 10 //then someone\ else can actually start organizing the report 11 and then you can just comment on it and say oh well 12 Paula: /that’s right yes\\ 13 Quentin: is that # what do you think is that okay While clearly enacting leadership by providing direction, the transactional orientation of the discourse is particularly evident in Quentin’s specific suggestions for ways of assisting and accelerating progress on the project, and getting value out of the time (line 2). He suggests bringing in professionals and making use of their ideas on the basis of a discussion that we actually record (line 3–4), as well as using another person to organize the report (line 10), leaving Paula free to comment on it (line 11). Quentin also attends to the relational dimension of workplace talk. His consistently consultative style is evident in the final line of example 1, where he 193

HylandK_Chap12_Finals.indd 193

12/1/2010 8:05:55 AM

Bloomsbury Companion to Discourse Analysis

checks that Paula is comfortable with his suggestions: what do you think is that okay (line 13). Example 2 provides further evidence as Quentin explains to one of his team members, Renee, how to complete a task for which she is responsible. Example 2 Context: meeting between Quentin and Renee, a junior member of his team. Quentin is describing to her how to do a new administrative task for which she is responsible. 1 Quentin: you know what the guidelines are eh 2 Renee: mhm 3 Quentin: we do that at the same time 4 Renee: mhm 5 Quentin: cos this seems to be + this is the first one + 6 in terms of + from your point of view eh 7 of understanding //eh yeah entitlement eh\ 8 Renee: /yeah yeah yeah yeah\\ When Renee does not immediately grasp what needs to be done, Quentin patiently repeats aspects of the task more explicitly and in greater detail. He says that he realizes this is the first time she has been required to understand this process from a different perspective, namely as the person responsible for administering it (lines 5–7). He provides encouragement and indicates a sympathetic rapport, by checking her understanding throughout, using a range of pragmatic particles: okay? you understand?, you see what I mean?, and eh?, a distinctively New Zealand pragmatic particle, strongly associated with Māori ethnicity. The particle eh (lines 1, 6, 7), in particular, functions to reduce formality and construct rapport. Quentin here enacts a responsive and patient leadermentor, helping Renee to understand her new task. Quentin also regularly engages in small talk and jokes with his team (classic relational behaviour), especially before and after team meetings. In example 3, he jokes that another colleague and he want to be additional boyfriends to another employee’s girlfriend. The girlfriend’s mother brings in lunch for the employee every day and they think it would be nice to get lunch delivered for them too. The team respond positively, and the humour contributes to group cohesion and builds solidarity. Example 3 Context: regular weekly meeting of Quentin’s whole team of seven. 1 Quentin: Jason gets lunch delivered by his girlfriend’s mother ... 2 so Rangi and I are thinking 3 maybe we should be extra boyfriends

194

HylandK_Chap12_Finals.indd 194

12/1/2010 8:05:55 AM

Discourse and the Workplace

Quentin’s interactions illustrate another feature of workplace meetings, discussed above, namely the different ways in which meetings are handled in different communities of practice, and especially by different cultural groups. The CEO of this Māori workplace commented to us that Quentin brought ‘a sense of ceremony’ to events. This was evident in a number of ways. When people came to visit the organization or when he went out to visit other organizations, Quentin used appropriate formal Māori protocol which paid explicit verbal respect to all those involved. He also consistently greeted and welcomed people formally to any meeting, and he opened and usually closed meetings with a formal, structured mihi (‘greeting’), paying attention to who was present, their affiliations, their relatives alive and dead, and the kaupapa (‘agenda’) of the meeting (Holmes et al. 2008). By contrast, meetings in Pākehā3 organizations typically began with a very brief and informal declaration such as ‘okay let’s get started’. Quentin’s enactment of his ethnic identity in his Māori workplace is thus deeply entwined with his role as an undoubted leader in matters relating to Māori culture or tikanga. He was clearly identified as the cultural leader by all those who worked with him. His proficiency in the Māori language also meant he could speak for the organization in Māori contexts when use of Māori was appropriate. Like a traditional Māori rangatira (chief), typically a male role, Quentin ‘weaves people together’ in his workplace, paying attention to their spiritual, relational and material needs. On the other hand, Māori leaders are also expected to be humble. The important Māori concept of whakaiti, that is appropriate modesty and humility, is fundamental to understanding the complex ways in which leadership is constructed, especially in the context of large meetings of workplace members. The analysis of Quentin’s discourse identified many ways in which he emphasized collegiality and reduced status and power differences, such as the use self-deprecating humour, and the extensive use of the colloquial pragmatic tag eh, mentioned above, a form that clearly expresses collegiality and emphasizes informality, as well as being strongly associated with Māori ethnicity. Such strategies enabled Quentin to avoid behaving in a way that could be interpreted as unacceptably proud, self-promoting or boastful, and to enact leadership and team membership in ways consistent with Māori cultural attitudes, values and norms. As mentioned above, there is a much greater use of the Māori language in this workplace than in most New Zealand workplaces, and this is especially true of interactions involving Quentin and his team. Quentin is a fluent Māori speaker and his team members all understand Māori. This means he has an additional resource for enacting ethnicity in workplace interactions. He and his team members frequently code-switched between Māori and

195

HylandK_Chap12_Finals.indd 195

12/1/2010 8:05:55 AM

Bloomsbury Companion to Discourse Analysis

English during their meetings, with whole turns often in Māori from some team members (Holmes et al. forthcoming). This is important background for understanding why example 4 is regarded as so amusing by Quentin’s team members. Example 4 Context: regular weekly meeting of Quentin’s team. Quentin is checking their future commitments 1 Quentin: we are at the [technician’s office] through this weekend 2 have I got that day right for you to see Paula 3 Paula: yeah the it’s the Tuesday //the\ 4 Quentin: /yeah\\ yeah 5 Paula: not Wednes- you had Tuesday the twenty second 6 Quentin: yeah I know Matariki pointed that out to me 7 and it’s my it’s my calendar 8 I still don’t know those days 9 the Māori names for days of the week 10 [general laughter] 11 so I look up and it’s the third and I think 12 oh let’s move In this example, Quentin acknowledges that he does not know the (new official) Māori names for the days of the week. To interpret accurately the significance of this admission, it is important to know that fluent Māori speakers such as Quentin learned a version of the days of the week which were transliterations from English (e.g. Mane, Tūrei, Wenerei). In recent years, the Māori Language Commission has been promoting new names for the days of the week, based on diverse and complex sources, both Māori and European (e.g. Rāhina, Rātū, Rāapa), and these are now the ‘official’ versions for use in formal contexts. Thus older fluent Māori speakers may find themselves wrongfooted by newer learners who are familiar with the currently prescribed ‘official’ names. Hence Quentin is here being self-deprecating in line with Māori values relating to acknowledging fault and being modest. But, ironically, the root cause of the need for modesty is in fact his highly valued and respected expertise and fluent ability in te reo Māori. Indeed the laughter that his admission occasions derives precisely from the fact that his team appreciate his outstanding linguistic skills and the humorous irony of his admission. In a variety of discursive ways then, in both one-to-one interactions and larger meetings, Quentin effectively addresses his organization’s distinctive transactional objectives as well as attending to the relational dimension of workplace interaction. Furthermore, he effectively accomplishes leadership through his talk, doing power in ways which enact his Māori identity, while 196

HylandK_Chap12_Finals.indd 196

12/1/2010 8:05:55 AM

Discourse and the Workplace

also emphasizing solidarity in ways compatible with the Māori values which imbue his ethnicized workplace.

New Directions in Workplace Discourse Research The analysis of workplace discourse has developed considerably in terms of data collection sites, methodology and approaches to analysis over the last 30 years. It seems likely that there will continue to be developments in all three areas in the next twenty years. In the last decade, workplace sites have expanded to include a much wider range of professional white-collar contexts, including government organizations and private companies, with large group meetings a common focus. It seems likely that there will be developments which will encompass data collection in small meetings, as well as perhaps more of the small talk and social interaction which takes place at work. As will be apparent, few researchers have ventured into blue collar worksites; they tend to be noisy and dirty and often rather uncomfortable places for academics undertaking research (but see Baxter and Wallace 2009; Goldstein 1997; Holmes and Stubbe 2003; Sunaoshi 2005). Nonetheless, this is undoubtedly another direction in which it is important to expand workplace discourse research. Similarly, there is little evidence other than hearsay on the interactional norms of skilled workers such as plumbers, painters and electricians. This also suggests that the definition of what counts as a workplace would benefit from further reflection, perhaps expanding to encompass more mobile sites such as the workspaces of professional sports teams (Wilson forthcoming) and travelling salespeople. Research on workplace discourse will inevitably extend to embrace ways of collecting more multimodal data and corresponding multimodal analysis (Jewitt 2009). Already some researchers have devised satisfactory ways of video-recording a range of workplace interactions. As technology improves and data collection devices become less intrusive, this will become standard. Correspondingly, workplace research could fruitfully develop in more interdisciplinary directions, with greater collaboration between scholars in discourse analysis and other disciplines such as business communication, leadership and media studies, and communication. Again there is some evidence of this already but there is scope for considerably more. And finally, there is scope for discourse analysts to develop greater reflexivity in their research methodologies (Sarangi and Candlin 2003), and an increased sensitivity to the needs and expectations of those they work with. In sum, there is much exciting work ahead, extending directions already apparent and developing new and innovative ways of analysing workplace discourse. 197

HylandK_Chap12_Finals.indd 197

12/1/2010 8:05:55 AM

Bloomsbury Companion to Discourse Analysis

Transcription Key yes Underlining indicates emphatic stress [laughs] : : Paralinguistic features and other information in square brackets, colons indicate start/finish + Pause of up to one second ... //......\ ... Simultaneous speech .../.......\\ ... Incomplete or cut-off utterance ... ... Section of transcript omitted # Signals end of “sentence” where it is ambiguous on paper All names are pseudonyms

Notes 1. I would like to express appreciation to Sharon Marsden and Meredith Marra who assisted with the literature review on which this chapter is based. 2. See Stubbe, Lane, Hilder, E. Vine, B. Vine, Marra, Holmes and Weatherall (2003) for an illustration of the results of a range of analytical approaches applied to one excerpt of workplace discourse. Sarangi and Roberts (1999) also include very valuable discussion of methodological issues. 3. Pākehā is a Māori word used to refer to New Zealanders of European (usually British) origin.

Key Readings Baxter, J. A. (2008), ‘Is it all tough talking at the top? A post-structuralist analysis of the construction of gendered speaker identities of British business leaders within interview narratives’, Gender and Language, 2(2), 197–222. Boxer, D. (2002), Applying Sociolinguistics. Amsterdam: John Benjamins. Chapter 6 on workplace discourse. Holmes, J. and Stubbe, M. (2003), Power and Politeness in the Workplace: A Sociolinguistic Analysis of Talk at Work. London: Longman. Koester, A. (2006), Investigating Workplace Discourse. London: Routledge. Mills, S. and Beeching, K. (eds) (2006), Journal of Politeness Research 2(1). Special issue. Politeness at Work. Spencer-Oatey, H. and Xing, J. (2008), ‘Issues of face in a Chinese business visit to Britain’, in H. Spencer-Oatey (ed.), Culturally Speaking: Managing Rapport through Talk. London: Continuum, pp. 391–416.

198

HylandK_Chap12_Finals.indd 198

12/1/2010 8:05:55 AM

13

Discourse and Gender Paul Baker

Chapter Overview Introduction The Turn to Discourse Discourse-based Approaches in Gender and Language Research A Sample Study: Cougars Future Directions Notes Key Readings

199 200 201 203 209 211 212

Introduction Robin Lakoff’s 1975 book Language and Women’s Place perhaps marks the start of a coherent field of language and gender research, although earlier, more isolated tracts like Jespersen’s (1922) chapter on ‘The Woman’, are sometimes used as shocking-hilarious examples of conceived wisdom of the time, a sexist view of women’s language use as deficient to men’s. Lakoff’s central argument was that men use language to dominate women. However, she acknowledged that ‘the data on which I am basing my claims have been gathered mainly by introspection . . . is the educated, white, middle-class group that the writer of this book identifies with less worthy of study than any other?’ (Lakoff 1975: 40). Unsurprisingly, the research was criticized for making over-generalizations and extrapolating conclusions based on observing a small sample of the researcher’s peers. Despite this, or perhaps because of this, her book can be credited for inspiring others to carry out further investigations, seeking to confirm, discredit or improve on her views. A later position advocated by popular self-help writer and interactional sociolinguist Deborah Tannen (1990), that men and women use language differently (whether due to socialization or other reasons), avoided accusing men

199

HylandK_Chap13_Finals.indd 199

12/1/2010 8:06:17 AM

Bloomsbury Companion to Discourse Analysis

of being bullies and women of being victims, but fell open to criticisms of an apolitical perspective. However, all three positions: deficit, dominance and difference are founded on the same over-generalizing principle – that males and females are different from each other. These approaches have tended to focus the area of language and gender around questions about usage (e.g. do men say word ‘x’ more than women) or about changing language to make it less sexist (see Kramer et al. 1978: 638). The very nature of such questions forces one to think in terms of differences and large binary categories that are made to appear fixed. Since the 1990s, a position which takes into account diversity, considers how particular women and men use language in specific settings, and the complex ways that gender interacts with other identity categories (Eckert and McConnnell-Ginet 1992) has arisen. Such an approach has helped to formulate an alternative set of research questions, which focus more on how language use helps to create, reflect and challenge discourses around gender. It could be argued that discourse has been the most influential concept in the area of language and gender in the past 20 years, altering the field beyond recognition from its beginnings in the 1970s. This chapter begins by outlining the ways that discourse has permeated language and gender research, giving an overview of recent research. It then moves on to give a sample study, based around analysing discourses surrounding the concept of the ‘cougar’, an emerging identity category used to describe women who have younger male partners. Finally, the chapter briefly addresses potential ways that the focus on discourse could help to develop the field of language and gender.

The Turn to Discourse As mentioned above, studies which have discourse as a central concept have become extremely popular in the field of language and gender over the past 20 years. As Gill (1995: 166) describes, language became a central concern across the social scientists, due to the ‘influence of post-structuralist ideas which stressed the thoroughly discursive, textual nature of social life’. Cameron (1998: 947) notes that in fact, this ‘linguistic’ turn was mainly a turn to discourse analysis. An illustration of the ‘turn’ can be demonstrated from an analysis of the 25 papers published in issues 1.1–3.1 (between 2007 and 2009) of the relatively new journal Gender and Language. Twenty (80%) of these papers mention the word discourse or its plural (not including examples in the references section of the articles). For those papers which did refer to discourse(s), the average number of citations across them was 25 mentions of the word per paper. Cameron (1998: 947) notes, however that ‘The popularity of this approach has produced competing varieties of feminist discourse analysis, based on what seem at times to be incommensurable assumptions and definitions 200

HylandK_Chap13_Finals.indd 200

12/1/2010 8:06:17 AM

Discourse and Gender

of the term discourse.’ The book Gender and Language Research Methodologies (Harrington et al. 2008) showcases the main current approaches in the field. Of its seven sections, three directly appear to address discourse: discursive psychology (DP), critical discourse analysis (CDA) and feminist poststructuralist discourse analysis (FPDA). However, orientation to the discourse approach is present in the other sections of the book. For example, in my chapter on corpus linguistics approaches, I show how large reference corpora can be exploited in order to uncover gendered discourses surrounding the terms bachelor and spinster (Baker 2008a), while Leap’s (2008) chapter in the section on queer theory analyses different discursive constructions of gay masculinity and Mullany’s (2008) chapter which takes an ethnographic, sociolinguistic approach, examines manager’s discourse in workplace settings. The above paragraph should help to demonstrate the different conceptualizations of discourse which Cameron refers to. Some gender and language studies have used discourse in a mainly neutral, descriptive way which approximates terms like genre, register or text type (see Gray and Biber this volume; Martin this volume). For example, Morgan (2007) writes of ‘women’s discourse’ while Motschenbacher (2007: 256) refers to ‘advertising discourse’. The former suggests that what is under study is the language (use and practices) of women, whereas the latter could mean language use occurring in the domain of advertising. However, Besnier (2007: 73) writes of ‘homophobic discourse’ inspired by American fundamentalist Christian organizations. Here we are still dealing with a type of language, but an explicit, critical judgement is made, and the implication is that we are not merely considering homophobic language in terms of words like ‘faggot’, but also the production of ‘meanings, metaphors, representations, images, stories, statements and so on that in some way together produce a particular version of events’ (Burr 1995: 48). Such language use may not actually use any pejorative terms and still count as ‘homophobic discourse’. More ambiguously, Zimman (2009: 75) refers to ‘coming out discourse’, which could indicate both the language and practices that occur when someone ‘comes out’ (discloses their sexuality), or the language and practices that surround the topic of coming out.

Discourse-based Approaches in Gender and Language Research One of the ways that discourse tends to be referred to most frequently in current writing in the field is via the term ‘gendered discourse’. This is sometimes, but not always linked to Jane Sunderland’s (2004) approach which is focused on identifying gendered discourses via traces in language use: People do not . . . recognise a discourse . . . in any straightforward way . . . Not only is it not identified or named, and is not self-evident or visible as a 201

HylandK_Chap13_Finals.indd 201

12/1/2010 8:06:17 AM

Bloomsbury Companion to Discourse Analysis

discrete chunk of a given text, it can never be ‘there’ in its entirety. What is there are certain linguistic features: ‘marks on a page’, words spoken or even people’s memories of previous conversations . . . which – if sufficient and coherent – may suggest that they are ‘traces’ of a particular discourse. (Sunderland 2004: 28) Sunderland’s approach moves discourse away from genres of language use, instead focusing on the concept of discourse as ways of representing the world. Such discourses, when identified, are named (although the naming is at the discretion of the namer, and as such, is a highly subjective process). For example, Sunderland (2004: 50) lists specific discourses such as ‘permissive discourse’, ‘equal opportunities discourse’ and ‘God’s will discourse’, noting also that discourses can be categorized in terms of their function, such as conservative, resistant, subversive or damaging. Another central facet of Sunderland’s approach is in identifying relationships between discourses, for example, pointing out that two discourses may be competing or mutually supportive or one may be dominant and the other subordinate. This relational aspect of discourse is one of the central concepts of Sunderland’s approach, helping to explain why people can appear to be inconsistent in their positions (they may be drawing on conflicting discourses) Related to Sunderland’s approach is FCDA (Feminist Critical Discourse Analysis), essentially a linking together of CDA and feminist linguistics, which is used perhaps more explicitly to critique ‘discourses which sustain a patriarchal social order: that is, relations of power that systematically privilege men as a social group and disadvantage, exclude and disempower women as a social group’ (Lazar 2005: 5). FCDA is thus concerned with taking the analytical tools developed in CDA in order to critique the ways that language use sustains unequal gender relations, for the purposes of emancipation and transformation. While FCDA also has the remit of showing how taken-for-granted assumptions around gender can be negotiated and contested as well as (re)produced, a third approach, offered by Judith Baxter, puts negotiation at its centre. Baxter’s FPDA ‘suggests that females always adopt multiple subject positions, and that it is far too reductive to constitute women in general, or indeed any individual woman, simply as victims of male oppression’ (Baxter 2003: 10). Instead, FPDA involves close, qualitative analyses of texts (often detailed transcripts of conversations) to show how participants (particularly those who may be conceived of as relatively powerless) can experience ‘moments of power’, whereas powerful people can be positioned as powerless. Baxter acknowledges that her poststructuralist approach, in rejecting ‘grand narratives’ and an overarching liberationist agenda, could be seen as nihilistic, although she argues that poststructuralist forms of analysis are well-placed to give voices to minority or oppressed groups, allowing them to be heard clearly alongside the more 202

HylandK_Chap13_Finals.indd 202

12/1/2010 8:06:18 AM

Discourse and Gender

dominant majority groups (2003: 37). All of the above approaches place emphasis on intertextuality, interdiscursivity and self-reflexivity, with the researcher needing to acknowledge their own theoretical positions and practices ‘lest these inadvertently contribute towards the perpetuation, rather than the subversion, of hierarchically differential treatment of women’ (Lazar 2005: 15). A further discourse-oriented approach (or set of approaches) to gender and language research is found in DP which combines a number of disciplines including conversation analysis (CA), ethnomethodology and rhetorical social psychology in order to critique ways that traditional psychological research understands concepts like attitudes, accounts and memory (Edwards and Potter 1992; Potter and Wetherell 1987). Through detailed analysis of transcripts, discursive psychologists show how speakers often produce inconsistent or conflicting versions or accounts. Some researchers working in gender and language have taken DP and introduced elements of poststructuralist theory or CDA, for example the work on young men’s talk about fatherhood by Edley and Wetherell (1999). Other researchers have shown how techniques used in CA can be adopted for feminist research (e.g. Kitzinger 2008), which although do not focus on discourse explicitly, are able to show how ‘gender – or sexuality, or power, or oppression – is produced and reproduced in interaction’ (Kitzinger 2008: 136). So while the approaches briefly described above may differ in terms of data collection and analysis, or underlying aims, they have all tended to place discourse as a central, if not the central component of their research.

A Sample Study: Cougars Considering the range of different discourse-analysis-based approaches to the field of gender and language, it is difficult to offer an example of analysis which is able to cover all perspectives. However, I have tried to draw on and combine various aspects from some of the approaches outlined above in order to examine the discursive construction of the cougar, which can be described as an emerging gendered identity. A cougar (sometimes called a puma, mountain lion or panther) is a wild, solitary mammal, native to the Americas. The term has recently begun to be used to describe women – an early example is by the Canadian sex and relationships columnist Valerie Gibson who wrote the relationship advice book Cougar: A Guide for Older Women Dating Younger Men in 2002. The concept seems to have become more widely known in 2007, when Mark Penn (Hilary Clinton’s chief strategist for her 2008 presidential campaign) released a book called Microtrends, which covers different emerging social groups in America, of which cougars was one. The term was further popularized in 2009 by an American situation comedy series called Cougar 203

HylandK_Chap13_Finals.indd 203

12/1/2010 8:06:18 AM

Bloomsbury Companion to Discourse Analysis

Town starring Courtney Cox who plays a divorced woman who embarks on relationships with younger men. The fact that such a group has been identified at all is of interest to researchers of gender and language – up until this point, there does not seem to have been a single popular term to refer to such women. There are many existing words which refer to older women in the English language, most of them ranging from patronizing to very offensive (crone, hag, biddy, old dear, bag, mare, witch etc.) although they rarely refer to such women in a sexually active sense (most of these nouns imply sexual unattractiveness). An explanation (although by no means a justification) for such terms is that society places a high price on a woman’s child-bearing potential, and once she is unable to have children, she is no longer viewed as desirable. Traditionally, relationships between younger women and older men have been seen as more usual or acceptable (with a term like sugar daddy to describe much older men). At her personal website (http://www.valeriegibson.com/book.htm – last accessed on 21 September 2010), Gibson defines a cougar in the following way: She’s the new breed of single, older woman – confident, sophisticated, desirable and sexy, she knows exactly what she wants. What she wants is younger men and lots of great sex. What she doesn’t want is children, cohabitation or commitment. It’s an irresistible combination for younger men who are more-than-willing to meet these sleek and sexy single women. So, if you’re a cougar (at any age) get ready to find out how to meet, catch and enjoy the perfect younger man and make them, and you, purr with pleasure! In many ways this description could be characterized as positive – cougars are described with the adjectives confident, sophisticated, desirable, sleek, irresistible and sexy (twice). The fact that cougars do not want children, cohabitation or commitment is not viewed by the author as problematic, but instead presented as part of the reason why such women are irresistible. Gibson, therefore, redefines the concept of the older woman who has relationships with younger men as an empowered identity, eschewing conservative discourses that women should want commitment and children. The book is written in a humorous style, and accordingly the term cougar also appears to be used somewhat humorously; for example, Gibson offers advice on how to ‘pounce back when [a relationship] ends’. An interesting aspect of emerging identity terms, however, is that they can often serve as the focal point for competing discourses. Consider, for example, the changes in meaning of the word gay. While the word was claimed by men and women who had same-sex relationships in the 1960s and 1970s as a positive identity term (to replace the more clinical-sounding homosexual), in the last decade or so, the word has undergone a further semantic shift, now being used 204

HylandK_Chap13_Finals.indd 204

12/1/2010 8:06:18 AM

Discourse and Gender

to refer to anything which is viewed as pathetic. This new use of gay is especially popular among young people, and while it could be argued that it is a separate meaning from the same-sex sexual identity category, this seems to be unlikely – rather, the ‘pathetic’ usage seems to have developed as a result of over-extension of gay being used with derogatory intent on people who experience same-sex desire. I would argue that the newer meaning of gay both reflects and propagates homophobic discourse, but it also indicates how simply inventing a new ‘positive’ term for a stigmatized group will not resolve that group’s problems if the underlying stigmatizing discourses are not also addressed. The term cougar is slightly different though, in that there is no existing popular negative word that it is replacing. I was thus interested in determining whether the positive cougar discourse which Gibson had developed would be accepted, or whether it would be contested, and if so, in what ways.

Locating the Data In order to do this I looked for texts which contained references to cougars from a range of different sources. These included large-scale corpora or archives of public texts (newspaper articles, transcripts of television chat shows, fiction etc) such as the COCA (Corpus of Contemporary American English), as well as the searchable online newspaper database LexisNexis. I also carried out searches of the internet in general, using Google, and I watched episodes of Cougar Town. All of these sources elicited numerous examples of people talking about cougars (or in some cases, cougars talking). In fact, many of the examples of cougar that I found did not refer to older women at all, but either to actual animals or other uses such as cars, sports teams, places or people. These examples told me something about some of the original meaning of cougars. For example, cougars as animals are often the subject of attacking verbs like ambush, attack, claw, corner, drag, entice, hunt, kill, loom, lunge, pounce, stalk, thrust. We are thus already primed with a set of associations about cougars (as predatory wild animals) when we hear the word, and these associations carry over when the word is used on older women.1 In total, I found 426 examples of cougar or cougars referring to women, from a wide variety of sources. My analysis then involved reading through these citations of cougar and their surrounding context, in order to identify how cougars were being represented via language use. I began to identify similarities between different types of representations, which enabled me to start to create a classification system. I initially simply classified the different representations in terms of whether the discourses in them were positive, negative or neutral, although unsurprisingly, there were cases where it was difficult to do this, as some citations could be interpreted as containing multiple and sometimes 205

HylandK_Chap13_Finals.indd 205

12/1/2010 8:06:18 AM

Bloomsbury Companion to Discourse Analysis

conflicting discourses. In addition, there is potentially more than one way that a discourse can be positive or negative. Rather than trying to quantify whether the overall discursive construction is positive or negative (with 426 cases it is difficult to draw conclusions, especially as multiple citations came from the same texts), instead, here I have chosen a single text which focuses on discussing the new phenomenon of cougars. This text was interesting to analyse as it contained a number of discourses which were found to be common elsewhere. For the purposes of providing a short example of discourse analysis, it is therefore a reasonably good representative text to use. The text is a spoken transcript that was found in the Corpus of Contemporary American English, taken from The Today Show on the American television channel NBC (2007). In the transcript, speakers 1 and 2 are female, while speaker 3 is male. 1 Speaker 1: All right, we’re going to move on now to today’s relationships. We’re talking about cougars. We’re not talking about the animals, we are talking about the 40 or 50-something women and the men they go after. Demi Moore, Kim Cattrall, and Jerry Hall; these are some of Hollywood’s wellknown cougars. So what’s in the cougar? Well, it’s not exactly what you think. 2 Speaker 2: Today’s cougar woman is someone who’s strong and independent. She’s intelligent, she’s witty, she’s fun, she’s charming. 3 Speaker 1: Cougar women aren’t just in Hollywood. More and more women in their 40s and 50s are challenging the way we think about older women. 4 Speaker 2: I’m constantly, constantly following the trends of haircuts, makeup, and clothing. Presentation is everything. I do not leave my house without stiletto heels, ever. 5 Speaker 1: Younger men are taking notice and going after the mature women. And some may wonder what the attraction is between an older woman and a younger man? 6 Speaker 3: I think for some women, they want a lot out of life. Men their age are seeming tired, maybe very stuck in their ways, not very flexible. So they’re looking to younger men to sort of inject more energy, more excitement. 7 Speaker 2: It’s not that I’m not attracted to men my own age either, it’s just that they’re not the ones who are approaching me. And it seems that the younger men are actually pursuing me. This short example (actually only the start of the full segment on cougars) helps to demonstrate some of the different ways that discourse analysis could be carried out on a text where gender plays a central role. 206

HylandK_Chap13_Finals.indd 206

12/1/2010 8:06:18 AM

Discourse and Gender

Discourse Analysis of Cougar: Different Perspectives An initial analysis might want to focus on linguistic markers in the text, particularly asking a question like – how are cougars represented via language? Most readers would probably immediately notice the string of positive-sounding adjectives that the presenter uses to describe cougars in line 2 (strong, independent, intelligent, witty, fun, charming). This is coupled with the assertion that cougar women are on the increase, with the implicature that because this is a popular phenomenon (and one initially associated with glamorous, trend-setting Hollywood actresses) that it must be a good one (line 3). The overall tone of the piece on cougars is light-hearted, with the presenters appearing interested at outlining this new phenomenon. In this sense then, there appears to be a positive discourse surrounding cougars and what they represent. The discussion also acknowledges, less explicitly, that the positive discourses they are creating around cougars are in opposition to something else, possibly more negative. In line 1, the interviewer asks ‘what’s in the cougar’ and answers her own question ‘Well, it’s not exactly what you think.’ And in line 3, the same speaker refers to how ‘women in their 40s and 50s are challenging the way we think about older women’. Here there is a presupposition that there is a way we have of thinking about older women, which is implied to need challenging. Thus, one of the characteristics of the positive discourse of cougars is that it is constructed as already oppositional to a negative discourse about older women, the implication being that older woman have not normally been characterized as strong, independent, fun and so on. Such a discourse is thus positive (even if there is an interdiscursive reference to the older negative discourse). However, it could be argued that as the interview continues, the discourse moves away from a view of cougars as strong and independent, and instead (perhaps inadvertently) draws on a third discourse – which positions the cougar as obsessed with physical appearance and fashion ‘presentation is everything’ (line 4). In this sense, it is by dressing in trendy clothes, wearing makeup and stiletto heels that the cougar becomes empowered, which results in them being noticed by younger men (line 5). From a feminist perspective, this could be argued as being problematic. The cougar is implied to be a sexual competitor (with younger women), but also a somewhat passive one – note how it is ‘younger men’ who are shown to be agents in line 5. They are the ones who are ‘taking notice and going after’ the ‘mature women’ who take the more passive object position. While the television piece thus accesses what could be called a ‘sex-positive discourse’ and acknowledges female desire, at the same time it positions cougars as primarily interested in attracting younger men and needing to dress in the latest fashions and high heels in order to do so. The text also positions men in different ways. Younger men are constructed (lines 5–6) as desirable, whereas older men are described in opposition more negatively 207

HylandK_Chap13_Finals.indd 207

12/1/2010 8:06:19 AM

Bloomsbury Companion to Discourse Analysis

as ‘tired, stuck in their ways, not very flexible’. Thus while women are able to escape the problems associated with old age (primarily by altering their appearance), the later part of the interview seems to confirm aspects of the original older discourse – that youth is attractive and being old is not. Interestingly then, while this text appears on the surface to be offering an empowered view of older women, there are some conflicting or ambivalent aspects to it. What aspects of this text would the different forms of discourse analysis described above focus on, and what conclusions would they reach? A feminist critical discourse analyst might emphasize the subtle gender hierarchy that seems to be implied in the text: young men appear to be placed at the top as they are viewed as physically desirable but also the ones who have the agency to pursue others. Cougars, while framed as attractive, are somewhat further down the hierarchy – they must work at being desirable by following fashions and even then, become objects of desire rather than the more active participants (this construction seems to be at odds with the original meaning of cougars as predatory wild animals). Finally, older men are at the bottom of the hierarchy, being constructed as unacceptable partners. Interestingly, younger women are not referred to explicitly. The fact that this hierarchy appears to be constructed by the female presenters, of which one appears to identify implicitly at least as a cougar herself (line 4), could be viewed as a form of complicity in terms of Connell’s (1995) framework of hegemonic masculinity. On the other hand, an FPDA might focus on the range of ways that cougars are positioned in the text, noting that some are empowering while others are not. One aspect of FPDA is its reflexivity; which could involve getting the analyst to ask a range of other people to conduct a discourse analysis of the same text, and noting the different possible interpretations. The term cougar, originating in North America, is clearly a very Western conceptualization of female sexuality. What would discourse analysts from other cultures notice about the text, which I (as a gay, male, middle-class, liberal, 30-something Westerner) may have missed? Other analysts may, for example, focus on the heteronormative aspects of the text, which in its discussion of male–female relationships, seems to ignore the existence of same-sex desire and how that may relate to older women. For example, in line 6, the speaker seems to imply that women only ever consider men as partners. Alternatively, taking Sunderland’s gendered discourses approach, we might want to focus on giving specific names to the gendered discourses that the text seems to refer to, for example, ‘older women can be attractive’, ‘older men as tired/inflexible’, ‘men chase, women allure’, ‘women must dress to please’, noting the relationships or conflicts between such discourses and how they may relate to more general ‘higher-order’ gendered discourses (such as ‘gender differences’) or potential non-gendered discourses in the text such as ‘youth is desirable’ or ‘Hollywood leads society’. 208

HylandK_Chap13_Finals.indd 208

12/1/2010 8:06:19 AM

Discourse and Gender

This would only be the beginning of any analysis, however. Only a small part of the segment on cougars is included here – the full transcript would need to be considered,2 along with the way that this item related to other items in the general structure of the television programme. Is this item positioned as the ‘light-hearted’ segment, suggesting that this is not a topic that we should be expected to take too seriously? Also, who is the typical audience of this segment? Generally, ‘relationship’ items tend to be aimed at women, and thus help to reinforce an expectation that women should be interested in relationships (and subsequently men should be less interested). A full analysis would need to consider other texts which discussed cougars in order to determine whether the combination of discourses here was also found in other contexts. The fact that this text occurred on daytime American television places restrictions on what can and cannot be said due to the diverse public audience. On a campus internet forum where the participants were young, male and relatively anonymous, I found cougars being discussed in much more explicitly sexualized (although also passivized) terms: ‘I’m gonna fuck the shit out of it I want me some cougar tail.’ Another strand of analysis could involve considering more closely the origins of the term cougar, and other associations around it. Valerie Gibson claims to have traced it back to bars in Vancouver (a city where cougars in their animal form are found in the surrounding mountains). Apparently the term was used by young heterosexual men: ‘as a put-down for older women . . . who would go home with whoever was left at the end of the night’, and it was only later that it was picked up by older women who engaged in a process of reclaiming. (From An ABC News item called ‘Are more younger women with older men’ (5 May 2005). See http://abcnews.go.com/Primetime/Health/story?id=731599 (last accessed on 21 September 2010). From the onset then, the discourses surrounding the term have been contested, reflecting a continuing ambivalence that society has with female agency, female desire and aging (particularly for women).

Future Directions The field of gender and language is always changing, often in response to the ways that societies frame gender relations. It could be argued that earlier feminist critiques of sexist language use have helped to create a climate where people are more sensitive regarding issues of gender inclusivity and representation. On the other hand, sexist discourses have not vanished – instead, at 209

HylandK_Chap13_Finals.indd 209

12/1/2010 8:06:19 AM

Bloomsbury Companion to Discourse Analysis

times they have simply become more sophisticated and thus more difficult to attack. Mills (1998: 247–8) argues, If texts are overtly sexist, they are easier to deal with, since overt sexism is now very easy to identify . . . It is clear that feminist pressure around the issue of sexism has had a major effect on the production and reception of texts. Sexism has not been eradicated but its nature has been transformed into this more indirect form of sexism. What is necessary now is a form of feminist analysis which can analyse the complexity of sexism . . . now that feminism has made sexism more problematic. Along similar lines, Lazar (2005: 17) argues, ‘The discourse of post-feminism is in urgent need of critique for it lulls one into thinking that struggles over the social transformation of the gender order have become defunct in the present time.’ Approaches to gender and language which use discourse analysis are ideally positioned to carry out such critiques because they are based on attention to detail as well as an awareness of the larger context that a word, sentence or single text appears in. As sexist discourses grow more subtle, discourse-based approaches must also develop. It is clear that gender and language research has never forced its practitioners to hold a single shared opinion about the ideal way to do research or even about research goals. While most feminist researchers wish to inspire social change for the better, some take a position which focuses more on improving the situation for women or other groups who are viewed as disadvantaged. Others may view everyone as potentially limited in various ways by gendered discourses. Some research tends to be concerned with pointing out power inequalities; other research is more aligned with positioning everyone as potentially powerful or powerless at various points. Rather than attempting to predict what the next ten years will bring, I instead will try to summarize what the most recent trends in the field have been, as it is likely that they will influence the next decade, even if they do not result in continuations of such work. There have been a number of recent publications, for example, which have foregrounded the role of women in the workplace, particularly the talk of and discourse about women as managers (see, for example, Baxter 2006, 2010; Ford 2008; Koller 2004; Mullany 2007). Such research is important and timely, focusing on women in explicitly empowered positions and also bringing up questions about ‘female language use’. When women are put in a position of power (until relatively recently associated with men), do they adopt traditional ‘male’ ways of interacting or do they do something else? How can femininity be incorporated into management style, and what discourses are emerging around women managers? 210

HylandK_Chap13_Finals.indd 210

12/1/2010 8:06:19 AM

Discourse and Gender

A further trend involves research on gender and language in non-Western contexts. For example, in 2008 the British Association of Applied Linguistics Gender and Language Special Interest Group ran an event called ‘Language and Gender in African Contexts’. The event featured presentations on the language of marriage ceremonies in Botswana and the construction of gendered discourses in the Cameroonian parliament. Issue 2.1 of the journal Gender and Language focused on the construction of Japanese women’s language. Other nonWestern research includes Sadiqi (2003) who examined women’s language in Morocco, and Mallinson and Childs (2007) who examined language and identity practices of black women in Appalachia. A related form of research has considered gender and language within the context of globalization (see, for example, Besnier 2007). Using reflexivity, Lazar (2005: 19) notes the ‘marked inclusion’ in the ways that such research has been conceptualized in the recent past has used. She shows for example, that Western research tends to be constructed as unmarked (and thus mainstream) in gender and language conferences, whereas anything from other cultures is positioned and marked as such. Hopefully we are moving towards a position where this distinction is lost, and all gender and language research is formulated with the same degree of markedness. And as part of the now well-established move towards focusing on different types of gendered identities, rather than considering a simple male/female binary, some researchers have become to explore the relationship between gender and sexuality in more detail (Baker 2008b; Cameron and Kulick 2003, 2006; McIlvenny 2002; Morrish and Sauntson 2007). Finally, research on gendered language use and young people, particularly young girls and women has also been a productive area recently (Besag 2006, Bohn and Matsumoto 2008, Casteñada-Peña 2009, Kamada 2009, Pichler 2009). These recent forms of research show how far the field of language and gender has moved away from Lakoff’s observations of her peer group (mainly white, American, middle-class, heterosexual people in relationships) to consider a much broader definition of language and gender. Such a focus is surely welcome, indicating a field which is becoming more inclusive as well as growing in size and scope. What should also be clear from this chapter is the important role that the concept of discourse (in its somewhat numerous forms, but especially those which are critical and reflexive) will continue to play in the area of gender and language in the coming years.

Notes 1. The concepts of semantic preference, semantic prosody and discourse prosody are relevant here. See Louw (1993), Stubbs (2001). 2. Later in the programme, for example, the female participants describe the attraction of young men in ways which position cougars as somewhat more sexually

211

HylandK_Chap13_Finals.indd 211

12/1/2010 8:06:19 AM

Bloomsbury Companion to Discourse Analysis predatory and patronizing: ‘Have you looked at any [young men] recently? Besides the hard body, which is totally a perk, the passion, I mean, they’re still following their dreams . . . They’re a little eager, like puppies . . . They’re trainable.’

Key Readings Baxter, J. (2003), Positioning Gender in Discourse: A Feminist Methodology. Basingstoke: Palgrave Macmillan. Cameron, D. (1998), ‘Gender, language and discourse: a review essay’, Signs: Journal of Women in Culture and Society, 23(4), 945–73. Harrington, K., Litosseliti, L., Sauntson, H. and Sunderland, J. (eds) (2008), Gender and Language Research Methodologies. Basingstoke: Palgrave MacMillan. Lazar, M. (ed.) (2005), Feminist Critical Discourse Analysis: Gender, Power and Ideology in Discourse. Basingstoke: Palgrave. Sunderland, J. (2004), Gendered Discourses. London: Palgrave.

212

HylandK_Chap13_Finals.indd 212

12/1/2010 8:06:20 AM

14

Discourse and the News Martin Montgomery

Chapter Overview Introduction Defining News and News Discourse Critical Approaches to the Discourse of Newspapers: Ideologies at Work Discourse and Power: The Broadcast News Interview Beyond Considerations of Power and Ideology in News Discourse? Note Key Readings

213 214 215 224 225 227 227

Introduction If journalism is ‘the sense-making practice of modernity’, as Hartley suggests (1996: 29), then it is hardly surprising that it has attracted a great deal of attention from linguists. Beginning with the pioneering work of Van Dijk (1988a, b) and Fowler (1991), on the one hand, and with Heritage and Greatbach’s work (1991) on the news interview, on the other hand, we can discern two main traditions of study of news discourse: the first deals mostly with newspapers and the structure of news in written text; the second deals with the broadcast news interview as spoken discourse and as a form of social interaction. The first approach expresses a long-standing concern with newspapers as the embodiment of forms of ideology under late capitalism. The second approach has been particularly concerned with issues of power and control as they are reflected in the engagement between public figures and news organizations. In this chapter, both traditions of work will be considered before exploring the limits as well as the possibilities of their approaches to news as discourse.

213

HylandK_Chap14_Fianls.indd 213

12/1/2010 8:06:38 AM

Bloomsbury Companion to Discourse Analysis

Defining News and News Discourse We need at the outset, however, some definition of the nature of news whose discourse has attracted so much attention. From a sociological perspective and from media studies there is an extensive literature on the nature of news (see, for example, Schudson 2003; Tumber 1999); and this has gone to some length to define its thematic qualities. In brief, news deals typically with the most recent events of a public scale and importance. Its focus is often negative, concerning war, famine, accident and disaster – ‘bad news makes good news’ as the phrase has it. It favours the immediate, the concrete and the personal rather than the abstract and the complicated process. It needs to be culturally relevant and unambiguous in its import, and it often focuses on powerful or elite persons or groups or blocs. These qualities are known as ‘news values’ (Galtung and Ruge 1965), the nature of which both define some core qualities of mainstream news and help to describe the process whereby the undifferentiated stream of everyday happenings becomes crystallized into the content of newspapers and news programmes. These core thematic characteristics of the news are thrown into sharper relief by considering an instance of what might be called ‘anti-news’ drawn from a daily news programme, ‘PM’ on BBC Radio 4. This programme normally offers ‘an evening look at the day’s events’ by providing a mix of ‘hard news’ and commentary with some human-interest items. Thus, it features a recognizable news routine of headlines, news reports and interviews. For a time, however, it introduced an innovatory slot, called Your News, read by the presenter near the end of the programme, a typical example of which would be the following. My son is in Philadelphia for a gathering of fellow cult-members who may one day realize they’ve been duped into cutting themselves off from their families. I had to cancel the order for my new car as the manufacturers had no idea when it would be made. I had a lovely text conversation with my daughter, the first one we’ve had since she stopped talking to me four years ago; maybe we’ll speak soon. I completely lost it at the crematorium, but I managed to get back enough composure to deliver Mum’s eulogy. ‘Your News’ provides an engaging, almost ironic, counterpoint to the rest of the PM programme. As a series of news items they are personal and private; they are only loosely anchored in time and place; and their effect is cumulative, depending in part on the effects of juxtaposition. Overall, they serve as a marked contrast to the public emphases of the remainder of the programme. 214

HylandK_Chap14_Fianls.indd 214

12/1/2010 8:06:39 AM

Discourse and the News

In terms of discourse, the deictic centre of these items tends to be an unspecified but serial set of first persons (I), who are thematized as the actor, agent or experiencer of the reported events. Other participants (‘my daughter’, ‘Mum’, ‘a colleague’) tend to be identified by relation to the first person deictic zero point. By contrast, the deictic centre of mainstream news discourse is quite different. Broadly, its deictic zero point is indeterminately the news organization (for instance, CNN or the BBC), or even ‘the news’ itself as an abstract impersonal discursive institution. Hard news tends to avoid the first person, is precise in its specification of time, place and participant from the outset of any story or news item, and it seeks to present events as if recorded from beyond any one individual’s perspective. Thus, mainstream news has claims to offer on a regular basis, in the words of Ekstrom (2002), ‘reliable, neutral and current factual information that is important and valuable for citizens in a democracy’; or, as Schudson (2003) puts it, ‘information and commentary on contemporary affairs taken to be publicly important’. Nonetheless, while the routine properties of the news may be summarized in this fashion, from the critical perspective of media scholars and linguists, the practices that underpin the shaping of mainstream news have always remained questionable: the routine qualities and practices of the news amount to a limiting construction of reality. Indeed, from a critical perspective news values and news discourse are shaped by and help in their turn to shape the ideologies of the news.

Critical Approaches to the Discourse of Newspapers: Ideologies at Work Ideology may be defined in the general (and relatively neutral) sense of the common beliefs, assumptions and opinions of a determinate group; or it may be defined in the more narrow sense of specific frameworks of meaning that serve to underpin (and routinely disguise) relations of power in particular socio-historical circumstances. Work on news discourse in the critical tradition has tended to move between the two, aspiring to the second sense of the term but often resting upon the first. Analysing news discourse as ideology in newspapers has found a useful starting point in the way readerships assume that each title has its own, politically biased, editorial line. In the British context, for instance, the Daily Telegraph is assumed at the level of electoral politics to favour the Conservative Party and is generally thought to be socially conservative. It is likewise widely believed that the Daily Mail espouses a socially conservative agenda. The Guardian on the other hand is more likely to be liberal on social issues and to favour parties of the left rather than the right. Many of the early analyses of news discourse 215

HylandK_Chap14_Fianls.indd 215

12/1/2010 8:06:40 AM

Bloomsbury Companion to Discourse Analysis

in the press focused on using linguistics to throw these ideologies into relief. In this vein scholars working with Fowler at the University of East Anglia produced a series of analyses which showed how particular ideological positions were constructed in the press. The construction or enunciation of these contrasting positions pivoted around grammatical choices in nominalization, voice (active versus passive) and transitivity. Both voice and transitivity were at stake in the following example analysed by Trew (1979). Compare, for example, the opening lines of the following two reports from similar British newspapers on 2 June 1975, both describing the same event: The Times RIOTING BLACKS SHOT DEAD BY POLICE AS ANC LEADERS MEET Eleven Africans were shot dead and 15 wounded when Rhodesian police opened fire on a rioting crowd of about 2,000 in the African Highfield township of Salisbury this afternoon. The shooting was the climax of a day of some violence. The Guardian POLICE SHOOT 11 DEAD IN SALISBURY RIOT Riot police shot and killed 11 African demonstrators and wounded 15 others here today in the Highfield African township on the outskirts of Salisbury. The number of casualties was confirmed by the police. Disturbances had broken out ... There are some important differences in the selection of lexical items. The Times, for instance, has ‘RIOTING BLACKS’ where the Guardian refers to ‘African demonstrators’. The Times refers to ‘violence’ whereas the Guardian refers to ‘disturbances’. But probably the most significant differences emerge in the contrasting structures of the headline and lead from each paper. The Times uses the passive: RIOTING BLACKS SHOT DEAD BY POLICE affected process agent This places in the foreground not so much those who perform the action as those who are on the receiving end of it (described, incidentally, as ‘rioting’). The Guardian, on the other hand, uses the active construction: POLICE SHOOT 11 DEAD agent process affected This clearly emphasizes the agency behind the action. Indeed, the Guardian report generally makes no attempt to displace responsibility away from 216

HylandK_Chap14_Fianls.indd 216

12/1/2010 8:06:40 AM

Discourse and the News

the police. By contrast, the first line of the Times report is not only in the passive: Eleven Africans were shot dead and 15 wounded affected process & aff process the agent of the process is left unspecified in this clause, to be identified by implication in the next, when police opened fire on a rioting crowd agent process affected But here, although the police are clearly the agent in an active construction, it is one in which they ‘open fire on’, a process which is significantly more neutral as to its consequences than ‘shooting dead’. In ideological terms we might suggest that The Times filters the event through a lens in which public order and the rule of law takes precedence over legitimate protest. The Guardian, on the other hand, favours the rights of legitimate protest over the repressive apparatuses of the state, especially in cases where the latter are exercised on behalf of a powerful ruling minority. This kind of approach, therefore, uses close systematic analysis of the different linguistic choices of the two newspapers to display their contrasting ideological emphases. Further analyses of this type may be found in Conboy 2006; Montgomery 1985/2008; Richardson 2007 and Simpson 1993. The latter, for example, makes UK tabloid newspapers the focus of inquiry and suggests various ways in which, by tone, by mode of address and by cultural allusion, they construct particular versions of community for their readership.

Patriarchal Ideology in Newspaper Discourse The same focus on linguistic choices in the transitivity system of the clause may be broadened out from ideology in the political field, narrowly defined, to address questions regarding the ideological construction or representation of gender. Take, for instance, accounts in tabloid newspapers of crimes of violence by men against women. A study by Kate Clark (see Toolan 1992: 208–27) of reports carried by the Sun newspaper highlights particular kinds of linguistic patterning that tend to shift the blame from the (male) perpetrator to his (female) victim. These patterns operate at the level of both syntax and vocabulary. Vocabulary for the victim ranges from items that depict her in familial (and hence respectable) terms, such as ‘bride’, ‘wife’, ‘mum’, ‘housewife’, 217

HylandK_Chap14_Fianls.indd 217

12/1/2010 8:06:40 AM

Bloomsbury Companion to Discourse Analysis

‘daughter’, ‘mother of two’, to items that depict the victim in extra-familial terms, such as ‘call-girl’, ‘Lolita’, ‘prostitute’, ‘blonde divorcee’, with a range of more neutral (but age-related) terms in between, such as ‘schoolgirl’, ‘girl’, ‘young woman’. The opposite ends of this range of items amount to a contrast between depicting women as either in relationships which are socially approved (mother) or not (Lolita). The perpetrators, on the other hand, are described in terms ranging from the socially neutral (name, occupation) to the extra-human: ‘monster’, ‘fiend’, ‘beast’, ‘maniac’. Two types of contrasting account seem to grow out of the contrast in vocabulary: where the victims are coded neutrally, or as ‘respectable’, the perpetrator becomes ‘extra-human’, hence: DOUBLE MURDER MANIAC PROWLS CITY OF TERROR or FIEND STRANGLES ONLY CHILD, 7 Conversely, however, where the victim is coded as disreputable then the ‘extrahuman’ vocabulary is less evident. Compare, for example, FIEND STRANGLES ONLY CHILD, 7 with SEX-STARVED SQUADDIE STRANGLED BLONDE, 16 LOVE BAN BY TEENAGE WIFE

The patterning of vocabulary, therefore, tends to exonerate the male perpetrators. The syntax also contributes to this general tendency in ways similar to those discussed above in the reports of police shootings in southern Africa. It is not unusual for these accounts of male violence to draw upon the passive voice with its attendant possibilities for deleting the male agent of the action, as in the following headline and lead sentence: GIRL 7 MURDERED WHILE MUM DRANK AT PUB Little Nicola Spencer was strangled in her bedsit home – while her Mum was out drinking and playing pool in local pubs. The headline and the lead begin with the affected, the girl, and the only agent encoded in these two sentences is not the criminal but ‘mum’, agent of a process, ‘drank’ in the headline – a process further elaborated in the subsequent sentence. Clark’s analysis is not based upon contrasting accounts of the same event (unlike those of rioting above) but on an intuitive sense that these patterns of representation are recurring and that other patterns of representation could 218

HylandK_Chap14_Fianls.indd 218

12/1/2010 8:06:40 AM

Discourse and the News

have been adopted but have not. Furthermore, the overall tendency of these patterns is to make the victim and the crime salient rather than the perpetrator. In cases where the perpetrator is made salient it is usually within a supernaturally evil or subhumanly bestial paradigm – one which has wider currency in popular culture and which treats such violence as by definition exceptional and extraordinary. In reality, however, there is much evidence to indicate that violence against women is not exceptional in the way the beast/fiend paradigm might suggest. Instead, the patterns noted by Clark seem to deflect attention away from the sheer ordinariness of violence against women and from what might be a fuller reflection upon its causes. Clark’s analysis of crimes of violence against women in the Sun is a specialized case of a wider concern evident in the critical discourse analysis of newspapers with how they depict and represent subordinated social groups. Important further work in this vein may be found in Van Dijk’s Racism and the Press (1991), Reisigl and Wodak’s Discourse and Discrimination (2001) and Richardson’s (Mis)Representing Islam (2004).

Nominalization and Ideology, Lexis and the ‘War on Terror’ It is common for research into ideology at work in the news to draw (even if quite loosely) on Hallidayan, systemic functional linguistics, focusing in particular on the system of transitivity, as discussed above. Ideological analysis within this tradition, however, has also examined features of nominalization. Both Fowler et al. (1979) and Fowler (1991) commented on the ideological effects of expressions such as ‘Attack on Protesters’ or ‘U.S. home loans crisis’, where a process that could be rendered by a clause is reduced to a noun phrase. In news writing there are strict premiums on space or time; and journalists are expressly schooled in compressing material to fill predetermined slots such as headlines. Thus ‘U.S. home loans crisis’ provides a neat, shorthand way of referring to ‘a crisis brought about by U.S. banks making loans to home buyers who were then unable to repay them’. However, in linguistic-ideological terms, the choice of a noun phrase to replace a full clause also has the effect of turning a complex set of logico-temporal processes into a unified entity. It is not just that in this case a tortuous set of financial arrangements that unfolded over time becomes rendered as a single referent; the noun phrase, in its discursive context, presupposes the existence of the thing to which it refers. In short, the nominalization naturalizes what occurred as a process, removing it from time, causality and history in a way that seems to render it inevitable, while lending it a spurious facticity. The significance of nominalization as a process of ideological naturalization lends support to more detailed work on patterns of lexis in the news, drawing upon the techniques of corpus linguistics. The ideologically loaded expression, 219

HylandK_Chap14_Fianls.indd 219

12/1/2010 8:06:40 AM

Bloomsbury Companion to Discourse Analysis

the war on terror, provides a case in point. A corpus linguistics approach allows us to examine the ways in which the phrase was adopted in the print media and to track its subsequent patterns of occurrence. The results, however (see Montgomery, 2009), suggest that there is no simple mirroring in the press of the neo-liberal perspective that may underlie the first use of the expression. For one thing a corpus linguistic approach to the use of the expression war on terror shows that in the British context it is often cited within quote marks, as may be seen in the following examples from the Observer and the Telegraph, where one-third of the lexical strings are of this kind. War on terror was mistake says Patten [TE] or terrorists or AlQaeda. The “war on terror” is just an excuse. [TE] who has them. The phrase war on terror was used to justify Iraq [TO] Pakistan troops ‘lose faith in war on terror’ [TE] own countrymen in the US-led war on terror. The troops were ambushed at [TE] Musharraf “do more” in the war on terror. A law was enacted earlier[TE] for President George W Bush’s “war on terror”. [TE] Freed jihadis put Pakistan’s war on terror ‘back to square one’, say [TE] go it alone in the “war on terror”, or work more closely with[TE] the service of the war on terror. His forces deployed in the [TO] no option but to join the war on terror. Even if the Pakistani leader[TO] the real battleground of the war on terror is undoubtedly one of the [TO] Cheney abused his power in war on terror By Tim Shipman in Washington [TE] Gen Sir Mike Jackson condemns ‘war on terror’ [TE] break from Tony Blair’s approach to the “war on terror”. [TE]

In effect, from soon after 9/11 the expression was often treated as problematic and handled with a degree of critical distance, as not belonging to the newspaper’s own discourse (attributed, accordingly, to a named source such as George Bush or Tony Blair). Note, for example, “War on terror was mistake says Patten”; or “Gen Sir Mike Jackson condemns ‘war on terror’ ”. Overall, indeed, citations of war on terror tend in any case to be more common in opinion and comment sections of the broadsheet press (in leading articles, op-ed pieces, columns and readers’ letters) than in straight news reporting or the tabloid press. The term, therefore, tends to be handled from the outset in news discourse as a rhetorical fulcrum in argumentative prose rather than as a simple referring expressing. On the other hand, the item terror has a strong life in print journalism after 9/11 in contexts other than the phrase war on terror, most typically in a series of environments where it occurs as the premodifier in a nominal group or noun phrase such as terror attacks, terror suspects, terror groups, terror laws, terror plots, terror networks, terror legislation, terror raids, terror police. Indeed, if these phrases as a group are compared with the expression war on terror, it is found that they occur more frequently in UK broadsheets over the six year period after 9/11 (roughly in the proportion of 11:9). Purely in frequency terms, therefore, this group – though less eye-catching than war on terror – are more significant. And the meaning-load 220

HylandK_Chap14_Fianls.indd 220

12/1/2010 8:06:40 AM

Discourse and the News

of these expressions works in two contrasting ways. Consider a series such as terror police, terror raids, terror laws and terror bill; compare these with a series such as terror attacks, terror plots, terror suspects, terror groups and terror networks. As nominal expressions their construction is grammatically similar. The attribution of terror, however, works quite differently from one series to the other. In the case of the latter series, a terror attack, for example, is routinely interpreted as one that inflicts terror. Similarly, a terror group is routinely assumed to be one devoted to inflicting terror. In the case of the former series, however, terror police, terror laws and terror raids activate assumptions not of inflicting terror but of preventing it: thus, terror raids by terror police backed up by terror laws are not assumed to be in the service of terror but are assumed to be against it. In effect, constructions that seem formally identical on the surface (e.g. terror laws versus terror groups) may activate quite asymmetrical sets of assumptions to distinguish between two kinds of beings or entities – those devoted to terror and those against it. Every time these nominalizations occur (with terror as premodifier) they invoke and condense a fundamental premise of post-9/11 geopolitical rhetoric, rarely any longer enunciated but given graphic initial shape in George W. Bush’s ‘Every nation in every region now has a decision to make. Either you are with us, or you are with the terrorists’ (Address to Joint Session of Congress, 20 September 2001). Thus, each time that we as readers make contrasting sense of terror watchdog versus terror sponsor within the bipolar force-field of this rhetoric we become aligned with it in an interpretively complicit fashion. This approach to ideology in the press, drawing upon techniques from corpus linguists (Montgomery 2009; see also Morley and Bailey 2009), is rather different in character from the earlier analyses of voice, transitivity and nominalization. First, the incidence of a pattern is tracked over time through a large corpus of data. Second, interpretation of an expression is supported by attention to its collocational environment. And finally this approach does not assume a simple determining effect by the text on the reader but acknowledges the role of reader in helping to confirm the ideology of the text.

Narrative Discourse and Professional Ideology in the News Finally, in relation to ideology and the press another body of work examines the relationship between news discourse and professional ideology or practice. Bell’s work, for instance, on news discourse shows how narrative operates differently in the news story than in everyday life (Bell 1991). The classic studies by Labov (2001) and Labov and Waletzky (1967), had shown how story discourse in natural settings (typically face-to-face interaction) adopted a straightforward relation to chronology, in which events are presented in the 221

HylandK_Chap14_Fianls.indd 221

12/1/2010 8:06:41 AM

Bloomsbury Companion to Discourse Analysis

order in which they occurred. Narrative discourse in the news, however, does not follow the chronological order in the presentation of events. Instead, ‘frontloading’ of significant information in the news item leads to disruption of the timeline of the story. Consider, for instance, the following item from a Scottish newspaper. (The numbering of the clauses indicates a likely timeline for the reported events.) THERE’S BEEN A CONFRONTATION... POLICE CALLED TO TAGGART FRACAS They were scenes that were not in the script. (8) Filming of a new series of Taggart turned to real life drama last night when (7) police were called to the shoot location in Glasgow’s west end. (6) A row broke out after people in a tenement claimed that (3) film crews moved into a close without permission, (4) prompting one concerned resident to call the police. (6) Tempers flared as (5) the show’s stars Blythe Duff, who plays DC Jackie Reid, and Alex Norton, who plays DCI Matt Burke, filmed in one of the flats. (9) Paddy Lyons, of Southpark Avenue Owners’ Association, said there had been concern among some residents about the nature of the storyline. He claimed (2) the association refused to give permission to use or film in the close, in part because they did not want their address to be associated with scenes understood to involve the murder of a doctor in a flat being used to manufacture the drug ecstasy. (9) Mr Lyons said: ‘The residents did not spend money on our close for STV to link it with criminal activity. We just didn’t want to be a part of it.’ (10) Last night residents and the production firm were understood to be making moves towards an amicable solution. Thus, likely timeline for the reported events would be follows: there had been concern among some residents about the nature of the storyline [of the film] [The Owners’ Association] did not want their address to be associated with scenes [of criminal activity] [They] refused to give permission to film in their close film crews moved into a close without permission the show’s stars Blythe Duff, ... , and Alex Norton, ... , filmed in one of the flats 222

HylandK_Chap14_Fianls.indd 222

12/1/2010 8:06:41 AM

Discourse and the News

prompting one concerned resident to call the police A row broke out police were called to the shoot location in Glasgow’s west end Filming of a new series of Taggart turned to real life drama last night Mr Lyons [=one concerned resident] said: ‘The residents did not spend money on our close for STV to link it with criminal activity. We just didn’t want to be a part of it.’ Last night residents and the production firm were understood to be making moves towards an amicable solution. The reason why the presented order of events departs from their likely temporal sequence is best explained by reference to the concept of news values. The journalistic report is written so as to lead on the most newsworthy element in the narrative – that a row broke out and that the police were called (referred to as a ‘real-life drama’): these happen to be the most recent events; they are ‘bad news’ in the sense of conflict; they are culturally relevant to the readership (the paper is a Glasgow broadsheet and the popular television police series, Taggart, is set against the backdrop of Glasgow.) Condensed into the opening of the news item, therefore, are the most newsworthy elements of the story considered from a journalistic perspective and without regard to chronology. From the perspective of editing the newspaper, this has the advantage of relegating less newsworthy information to the end, thus simplifying the task of subediting copy to a required length. Known as ‘the inverted pyramid’, this is a structure which evolved over the course of the twentieth century as a specific and well-understood writing practice within the journalistic profession. Paradoxically, the evolution of the inverted pyramid in journalism with its refocusing of the event line to accentuate news values took place in step with a growing commitment to objectivity in journalistic reporting. The history of objectivity in journalism is traced by Schudson in part to the introduction of the telegraph and the subsequent growth of press agencies, who, in selling reports by wire to multiple outlets, aimed for a neutral style which would offend none of their partisan newspaper clients (Schudson 2003). Gradually the style moved from the wire to become generalized across the outlets themselves. Other factors, such as the disillusionment with propaganda by the end of the First World War, doubtless also played a part. But ultimately ‘objectivity’ becomes ‘a strategic ritual’ as the sociologist Tuchman says (1972, 1978) and an important professional tool-of-trade for the working journalist. White (1994), however, from a discourse analytic perspective, shows how the system of appraisal (see, for example, Martin and White 2005) can be used to 223

HylandK_Chap14_Fianls.indd 223

12/1/2010 8:06:41 AM

Bloomsbury Companion to Discourse Analysis

reflect upon this oft-cited professional commitment to objectivity and impartiality in reporting, and he calls into question whether objectivity is ever truly possible. [T]he strictly ‘objective’ text is constructed in such a way that there is no explicit linguistic evidence of the author’s value judgements. All value judgements are backgrounded or ‘naturalised’ in the sense that the way the event is construed is presented as the only way of talking about it. In this context, therefore, the ‘impartiality’ or the ‘factuality’ of a text are not measures of the degree to which it accurately reflects reality – as human subjects we use language to construct rather than reflect reality – but measures, rather, of the success of the text in presenting its underlying set of value judgements and ideologically informed responses as ‘natural’ and ‘normal’, as fact rather than opinion, as knowledge rather than belief. ‘Objectivity’, therefore, is an effect created through language (a ‘rhetorical’ effect) rather than a question of being ‘true to nature’. Overall, therefore, analysing ideology at work in newspapers suggests that the language of the press, even at its most apparently objective moments, gives us fundamentally a skewed, even flawed, representation of reality. Newspapers, from this perspective offer versions of the world that are inescapably ‘structured in dominance’, reflecting the interests of the state, of patriarchy, and of other powerful social groups. The task of discourse analysis is to reveal these processes of misrepresentation at work in all their close linguistic detail.

Discourse and Power: The Broadcast News Interview By comparison with the attention paid to newsprint, broadcast news discourse has been relatively neglected. Only the broadcast news interview has attracted significant numbers of researchers, prompted perhaps by the sense that it ‘is now one of the most widely used and extensively developed formats for public communication in the world’ (Corner 1999: 37; see also Ekstrom et al. 2006). As a broadcast event, the news interview has – especially more recently – been associated with the rise of the high-profile journalistic interviewer, who acts as a critical intermediary in the way public figures communicate with their public and who performs the role of the Fourth Estate in holding them to account. If newsprint has been seen as the site of ideological work in the service of power, the news interview has been studied as a way of ‘doing accountability’ on the powerful. This is the focus of a major study by Clayman and Heritage (2002). They note the fundamental character of the turn-taking machinery in the news 224

HylandK_Chap14_Fianls.indd 224

12/1/2010 8:06:41 AM

Discourse and the News

interview based upon an institutionally defined ‘turn-type pre-allocation’ in which ‘the activities of asking and answering questions are pre-allocated to the roles of interviewer and interviewee’. On this basis they characterize recurrent kinds of asking and answering behaviour in the news interview, recognizing for example that ‘there is a widespread perception that politicians are often evasive under questioning from members of the news media’ and detailing some of the techniques by which interviewees either fail to answer the question they have been asked or attempt to shift its agenda. Indeed they note that ‘it may be fairly claimed that the interactional accountability of answering questions is the fundamental basis for the public accountability of public figures’ (2002: 235). They also note, however, that ‘the impetus toward evasiveness is understandable in the context of the contemporary news interview, which is so often adversarial in character. Hostile questions, if answered straightforwardly, can inflict real damage on an interviewee’s policy objectives and career prospects’ (2002: 238). Indeed a substantial portion of their study examines the adversarial forms of questioning adopted by interviewers in which interviewers set the agenda and challenge the interviewee, sometimes in hostile and accusatory fashion. Clayman and Heritage’s account reveals the workings of the news interview as a particular kind of institutionalized and broadcast event. The news interview is talk for an overhearing audience in which the interviewer adopts the role of tribune of the people and plays ‘the devil’s advocate’ but also must operate within certain kinds of professional codes requiring the display of objectivity, impartiality and neutrality. Indeed the crux of their account revolves around the question ‘how do interviewers manage to assert themselves in an adversarial manner while maintaining a formally impartial or neutralistic posture?’ (151). They provide precise detail regarding how interviewers maintain this ‘formally impartial and neutralistic posture’. Interviewers, they demonstrate, often ‘speak on behalf of a third party’, introducing contentious views into the interview in the form of ‘third party attributed statements’ for the interviewee to agree or disagree with. Third party statements can be attributed to public figures (e.g. a rival politician), to accredited experts, or to ‘the public at large’ (‘there are people watching who will say ... ’). And it is noticeable that in responding to these, interviewees may well seek to refute them but in a way that does not challenge the interviewer’s neutralistic posture.

Beyond Considerations of Power and Ideology in News Discourse? While an impressive body of research has accumulated in both traditions, it has been subject to some criticism on the grounds that both approaches have 225

HylandK_Chap14_Fianls.indd 225

12/1/2010 8:06:42 AM

Bloomsbury Companion to Discourse Analysis

been limited in terms of selection of data and of text-type (see Stubbs 1996, 1997) and in terms of their underpinning perspective (see Widdowson 2004.) In the case of the critical discourse approach to newspapers, the analysis is shaped by the ideological concerns of the analyst – with, for instance, racism, sexism, the treatment of minorities and so on. It has been criticized, particularly by Widdowson, for selecting its data to prove its point (Widdowson 2004, chapter 6.). More seriously, perhaps, there is a tendency for it to associate a particular form with a particular function, whether this be agent deletion at the level of the clause associated with ideological suppression of the agency of key social actors, or nominalization associated with the naturalization of social processes (see discussion in Montgomery 1986). In respect of nominalization, Billig (2008) has shown how reductive it can be to assume a simple correlation of ideological role with the form.1 More generally, there has been a tendency to approach newspaper discourse as symptomatic of concerns defined elsewhere within a generally critical account of the role of the media in mass society. As a result analyses have tended to overlook the specific character of news discourse as displayed, for instance, in terms of the wide variety of text types or genres implicated in nearly all news outlets. Those analyses that have attended closely to the specificities of news discourse as reflected in the range of subgeneric types through which it is constituted (e.g. headline, editorial, op-ed piece, news report or readers’ letters) have generally been more persuasive (see Bell 1991; Richardson 2007; White 1994). The approach to broadcast news is also not immune to these criticisms. The overriding focus of much research on the adversarial political interview has far outweighed attention to other kinds of news interview, with the consequent neglect of other interview types such as: (1) the interview between journalists themselves, (though see Kroon 2009; Montgomery 2006); (2) the interview with a witness, survivor or bystander; or (3) the interview with an expert. In practice all of these are so routine in news broadcasting as to constitute clear generic subtypes of the broadcast news interview (see discussions in Montgomery 2007, 2008a, 2010) where the style, turn-types, structure and appeal to the broadcast audience vary considerably. And yet considerations of power are less obviously relevant to the study of these other types than in the study of the accountability interview with a public figure. In any case, analyses of the accountability interview run the risk of adopting a one-sided approach in which the analyst starts from the position of finding public figures evasive under questioning on air. An alternative approach would attempt to explicate the grounds under which interviews of this type produce a shared alignment between the journalist and the audience against the public figure (rather than simply replicate that alignment in the analysis) and then compare it with the way in which quite different kinds of alignment are favoured in witness, experiential and expert interviews. 226

HylandK_Chap14_Fianls.indd 226

12/1/2010 8:06:42 AM

Discourse and the News

Scannell (1998) commented on two different kinds of approach to mediated communication. It can be approached from the perspective of a hermeneutics of suspicion or a hermeneutics of trust. In everyday life our habitual attitude in communication rests upon a hermeneutics of trust. We assume that communication works – that it rests, for instance, upon the reciprocity of perspectives and the cooperative principle unless and until we run into communicative difficulties. Much work on news discourse, however, assumes the opposite – that we cannot trust what the news offers us as versions of the world. The habitual attitude of the everyday reader or viewer of the news is likely to be much more a mixture of the two – of both trust and suspicion. Work on news discourse needs to rediscover something of the ordinary systematics of communication, if it is to make sense of how news works as discourse. Future work on news discourse is likely to explore the generic differences between different kinds of news discourse in greater detail. News discourse, after all, is a mediated institutional discourse dependent for its production, for the most part, on large and costly organizational arrangements, using different kinds of media platform, and capable of reaching large publics only through significant technological innovations. The way news operates in broadcasting is not the same as in print, which in any case is facing significant competition from online journalism. And so we are faced with a range of different and changing discourses of evolving generic types. While there have been suggestive comments about the growing conversationalization of the news, perhaps the greatest challenge facing future work on news discourse is to understand the directions and the determinations of the ways in which it changes.

Note 1. Billig (2008) makes the rather telling point that critical discourse analysis itself is replete with nominalization. The term ‘nominalization’ itself would be a prime example. But see also ‘informalization’, ‘marketisation’, ‘conversationalisation’.

Key Readings Clayman, S. and Heritage, J. (2002), The News Interview: Journalists and Public Figures on the Air. Cambridge: Cambridge University Press. Montgomery, M. (2007), The Discourse of Broadcast News: a Linguistic Approach. London: Routledge. Richardson, J. E. (2007), Analysing Newspapers: An Approach from Critical Discourse Analysis. London: Palgrave Macmillan.

227

HylandK_Chap14_Fianls.indd 227

12/1/2010 8:06:42 AM

15

Discourse and Computer-mediated Communication Julia Davies

Chapter Overview Introduction The Online Context The New Literacy Studies The Fabric of the Text: Multimodality and Semiotic Resources Conclusions and Future Directions Notes Key Readings

228 229 229 230 242 242 243

Introduction This chapter focuses on the discourses and meanings produced when people communicate using computerized technologies. Specifically I concentrate on online discourse; exploring the fabric, nature and context of online texts and how these impact on meanings. While the internet contains billions of articles in uneditable format similar to paper-based texts, social networking sites (e.g. Facebook, MySpace, Flickr) comprise interactive, dynamic texts, where people meet, perform social acts and live their lives (Davies 2009; Davies and Merchant 2009). As the number of interactive online participants multiplies, further spaces will doubtless generate each with their own context-specific discursive conventions and practices for meaning-making. While the internet is a global arena, people’s online social networking commonly draws on existing local, geographically and socially defined communities. Sometimes however, interactants create new kinds of ‘online locality’, 228

HylandK_Chap15_Finals.indd 228

12/1/2010 8:07:02 AM

Discourse and Computer-mediated Communication

with their own conventions, social histories and linguistic rituals (Davies 2008). In this chapter I discuss examples from blogs, exemplifying social networking, textual presentations of the self and local communities within a global setting.

The Online Context Much of what we know about analysing discourse in offline contexts can be applied to online discourse, so that we might, for example, see how language is used to perform social tasks, like enacting friendship (Coates 1996); exerting power (Mayes 2010), or being polite (Watts 2003). When online, we inevitably draw on previous social knowledge of how to conduct ourselves, so analysis is likely to reveal that many of these social acts when performed online involve similar conventions to those found in offline spaces. We are also developing new rituals and ways of signifying meaning in online contexts, such as hyperlinking to show friendship, or using emoticons (smilies and such like) (Davies 2006) to symbolize how we feel. In analysing Computer Mediated Communication (CMC) therefore, we might attend to images, emoticons, sound files, font design and so on as these comprise the ‘semiotic resources’ of online texts. Sometimes we can trace in online texts, new routines and conventions that implement ‘old’ codes, like using an alphabetic grade to feedback responses after eBay transactions1 (Davies 2008); such features signal the emergence of new social practices. In this chapter, I focus on a particular type of CMC, namely blogs. I am interested in how individuals present themselves through text and connect with others and this suggests a focus on textual cohesion (Halliday and Hasan 1985) and on textual features and conventions that instantiate identity presentation. I therefore concentrate on looking at ways of making links and at presenting the self. I position my work within the field of The New Literacy Studies, seeing text-making and text-production as social practices.

The New Literacy Studies Rooted in sociolinguistics and language studies, and building on the idea that language is a social semiotic (Halliday 1985), the New Literacy Studies (NLS) regards reading, writing and meaning-making as situated within specific social practices and discourses (Street 2003). Being literate therefore equates to communicative competence (Hymes 1972) where interlocutors require all kinds of sociocultural knowledge in order to competently/confidently interact. Evocative of speech act theory (Searle 1969), which suggests words are 229

HylandK_Chap15_Finals.indd 229

12/1/2010 8:07:03 AM

Bloomsbury Companion to Discourse Analysis

performative, – they actually do things and can be performed like actions, making a difference in the world or even producing a different world – Barton and Hamilton (1998: 3) argue that, ‘Literacy is primarily something people do; it is an activity located in the space between thought and text’. Regarding literacy as a set of practices rather than a narrowly defined set of skills leads researchers to collect data and immerse themselves in settings where literacy practices occur, to explore those literacy practices in relation to contexts. My own research adopts this ethnographic spirit, involving observation and extrapolation of what seem to be socially significant cues and rituals. I describe how the internet allows people to develop new social practices and how it allows them to operate in new ways (Davies 2006a, b, 2007, 2008). The examples used in this chapter therefore represent patternings, routines, conventions and practices observed across many blogs.

The Fabric of the Text: Multimodality and Semiotic Resources This chapter looks at the ‘fabric’, ‘nature’ and ‘context’ of online texts; theorists increasingly argue that we need to take multimodal approaches when analysing interaction in face-to-face situations (Taylor 2006) and this pertains to literacy research too (Kress 2010). Jewitt (2005) describes the enhanced role of images, particularly in screen-based texts, where the visual, not just superficially embellishing, often plays a central semiotic role. Texts are never purely linguistic; Halliday and Hasan (1985) emphasize that a text is a semiotic resource embodying discourse and creating new meanings. Kress (2010: 79) explains, Mode is a socially shaped and culturally given semiotic resource for making meaning. Image, writing layout, music, gesture, speech, moving image, soundtrack and 3D objects are examples of modes used in representation and communication. The term ‘semiotic resource’ allows us to infer a broad spectrum of meaning making modes. Language is not always the primary carrier of meaning in online texts, – such as in a Facebook photo album, where a series of images might suggest a biographical story about a Facebook keeper. Moreover an image in one person’s Flickr stream2 may mean something different when it is copied and pasted to someone else’s blog post. Alternatively, that same image may import meanings to the blog embedded from its original Flickr context, since one can ‘trace back’ digitally, to explore the provenance of that image. Tracing back, or searching forward, means readers can overlay texts with meanings from each other, and these meanings can be influenced by the ‘route’ that a reader takes 230

HylandK_Chap15_Finals.indd 230

12/1/2010 8:07:03 AM

Discourse and Computer-mediated Communication

before arriving at any given instance of text. It is appropriate, therefore, to take a multimodal, context-aware approach when looking at online meaning-making; and the context may certainly exist beyond one specific online space.

D/discourses Gee (1996) distinguishes between ‘discourse’ and ‘Discourse’. Discourses with a large ‘D’ comprise social practices, mental entities and material realities. Gee talks of ‘Ways of behaving, interacting, valuing, thinking, believing, speaking . . . of being in the world’ (Gee 1996: viii). D/discourses are thus always and everywhere social, and products of social histories. I have described eBay for example, as a specific discursive space, comprising multimodal Discourses (Davies 2008); being recognized as a real ‘eBayer’ means following specific discursive practices that are part of the Discourse of eBay – this involves selling and buying in the space and following the rules, but further to these basic behaviours, there are other ways in which eBayers discursively situate themselves within the site. These are the ways that might be described as the culture, the socially accepted practices which may not be expressly articulated, but which nevertheless are multimodally codified practices and to understand them is to have ‘communicative competence’ (Hymes 1972) or to be ‘literate’ within eBay.

Blogs In this chapter, I describe the ‘performative’ aspect of blogs – how bloggers present themselves; secondly, I describe how bloggers link to other bloggers to form social-allegiances – in a process which Danath and boyd (2004: 2) refer to as ‘public displays of connection’. Cohesive devices make links within and across texts; through the production of texts which link with each other, individuals collaborate and weave a network of associations that are both explicitly articulated and implicitly expressed through textual cohesion. The cohesion operates on many levels, being related to structure (e.g. the grammar or layout), modes (e.g. images or written words) and also to semantics (or meanings). The composition of blogs and how they cohere allow particular kinds of textual self to be presented. Myers (2010: 77) suggests that ‘Blogs are defined by what they link to and how they link’; this complex configuration of blogs in a seemingly bounded network is recognized by the term ‘blogosphere’.

Recognizing a Blog The generally accepted definition is that blogs are regularly updated online dateordered texts, with most recent posts positioned at the ‘top’ of the screen. The 231

HylandK_Chap15_Finals.indd 231

12/1/2010 8:07:03 AM

Bloomsbury Companion to Discourse Analysis

term ‘blog’ developed from two words – web log, and this, as well as the dateordering, might imply that all blogs are online personal diaries. However, they are used for many purposes, often multiply-authored and not always reflecting on bloggers’ lives. With Merchant (Davies and Merchant 2009: 25), I identified some blog genres, including academic blogs, community-arts blogs, fan sites, through to citizen journalism and fictional blogs. These vary in purpose and content and demonstrate the range encapsulated within the category ‘blog’. Readers can access previous posts by clicking on dated links in the sidebar or can search for topics using tags, or categories, developed by the blogger to suit their focus (see Figure 15.1). Bloggers can allow comments to be made by readers, anonymously or non-anonymously; they can link to other bloggers via links in a blogroll, and/or subscribing via an RSS feed.3 Blogging templates usually include a space for a profile. Blogs can be made private, partly private, or totally public; most blogs are public.

The Local Expert Blog In the next section, I take a number of examples from one blog, ‘Combing my Hair’ using it to show a range of ways in which bloggers can use links and other strategies to perform a particular identity, in this case the ‘cycling expert’. Blood (2002) identifies two main types of blog; first the online diary type, with just a few links; the other being a kind of database of hyperlinks which have resulted Blog title Blog title Subscription invitation

post title

Profile

image (Taken from film website)

Blog post - text (including hyperlinks)

Author; Date and time of post

Tags & Categories

Blog roll

Archives

Figure 15.1 Example blog page (Gareth, 23 February 2010)

232

HylandK_Chap15_Finals.indd 232

12/1/2010 8:07:03 AM

Discourse and Computer-mediated Communication

from the blogger’s research. Clearly there is a continuum across these two main types, and bloggers may dabble across both extremes. As will be seen, ‘Combing my Hair’ is hybrid type of blog which diarizes the bloggers’ cycling life, but which also includes links to cycling and bike sites.

Presenting the Expert – Self as Expert Figure 15.1 provides an example of a typical blog layout and is drawn from the cycling expert blog, ‘Combing my hair’. All posts in this blog are consistently concerned with bikes or cycling; the categories include words like ‘cycling’, ‘clothing’, ‘hubcaps’ and so on; most of the blogs the author links to are also about cycling. Thus the sidebar itself authenticates the blogger as thoroughly immersed in bikes and cycling. The author not only self-identifies as a cyclist, he demonstrates that he is part of the cycling community by displaying links in his blogroll to others within that community (structural cohesion). The blogger consistently uses specialist language and insider knowledge about cycling and cycling culture (semantic cohesion).

The Profile – Saying and Showing the Self The blogger’s profile provides further, consistent cycling expert information. Underneath a large photograph of himself cycling up a hill, the blogger writes, I’m Gareth and this blog contains my musings on bicycles and other things that interest me when I am not thinking about art and education – having said that I am always on the look out for good cycling photographs and videos. [ . . . ] Oh, and I like climbing hills on my bike. In fact I usually like climbing them more than I like descending them. I seem to be quite good at it, possibly because one of my esteemed colleagues said I looked like a ‘lanky streak of . . . ’ (I am sure you know the rest – I took it as a compliment). The photograph and accompanying text allows Gareth to display he is an expert and an accomplished hill climber while also explicitly articulating this idea. Further, the blog is filled with professional-looking photographs of cycling, many close-ups, with careful, technical descriptions of bike parts and so on. Goffman (1969) makes the distinction between ‘giving off’ clues unconsciously, (e.g. through bearing, gestures, clothing, language) while also performing an identity, by purposefully adopting particular ways of behaving. These two dimensions for the presentation of self are not always congruent; it may thus be possible to come across as disingenuous if one performs a role in one way but ‘gives off’ an inconsistent presentation in other ways. 233

HylandK_Chap15_Finals.indd 233

12/1/2010 8:07:04 AM

Bloomsbury Companion to Discourse Analysis

Up Close and Personal We also see in the profile, a sense of friendship being extended. ‘I’m Gareth’, introduces the writer on first name, friendly terms. Perhaps it is ironic when he uses the exclamatory ‘oh and . . . ’ implying he forgot to say something – but of course writing can be smoothly edited, and so this is ‘as if’ spoken discourse. This sense of the jokey colloquial gives the blog a direct, informal style, where one feels that the writer is attempting to transcend the fact that the audience could be global, unfriendly or unsympathetic, this assumption shows trust, and there is a sense that we are being drawn into the local space of the blogger. Gareth’s profile suggests he is musical, with his blog-title referring to his favourite song by Jim White; this theme of musicality is, suggested in many posts through subtle motifs, a kind of hinterland, adding an additional dimension to the cycling expert identity. Gareth makes oblique references to songs, but rarely hyperlinks to them; he does not allow them to detract too much from the cycling topic. For example, a post ‘Reasons to be fearful’, obliquely references a song, by the late Ian Drury, ‘Reasons to be cheerful’ – through the phonological cohesion between the words ‘fearful’ and ‘cheerful’. There is a sense of underemphasized expert knowledge here; intertextual references are suggested, but rarely signalled – as so easily could be done – via hyperlinks to the songs mentioned. It almost seems that the blogger is writing as much for himself as for an audience to whom he needs to explain everything; we witness again the writer ‘giving off’ aspects of the self, possibly by design, that are consistent with what is articulated elsewhere.

Ventriloquism and Hyperlinks The text on the post shown in Figure 15.1 reads as follows: Monday saw Bike Night at the Sheffield Showroom. The main feature was Peter Yates’ 1979 coming of age movie Breaking Away. A heartwarming tale it also serves as a document of how road bikes and cycling position have changed over three decades. The film was followed by MC Spandex’s amusing critique of fixie culture, and then an extended film about Japanese fixie culture – we retired to the bar to discuss our approach to securing the team prize at the Brompton World Championships. (Gareth 23 February 2010) The 83-word paragraph presents the blogger (Gareth) as discursively situated within the Discourses of cycling and blogging. Gareth refers to a ‘Bike Night’ where cycling related films were shown; he mentions a Brompton (a foldingbike) world championship race; a ‘fixie culture’ – a subculture within cycling; and how ‘cycling position’ has changed over 30 years. Gareth thus concisely refers to four bike-related events or concepts. Further, the specialist terms ‘fixie culture’ and ‘Brompton World Championships’ are hyperlinks. The first links 234

HylandK_Chap15_Finals.indd 234

12/1/2010 8:07:04 AM

Discourse and Computer-mediated Communication

to a video in YouTube – the same film he saw in the cinema about ‘fixie culture’; the second links to details of the ‘Brompton World Championship’. Through this hyperlink referencing, and the ease with which he uses ‘insider language’ the blogger demonstrates his knowledge of bike culture and his ability to navigate to cycling related semiotic resources online. Expertise in locating relevant information is highly valued online, (Blood 2007; Gee 2004) and the links are embedded in the sentences without altering the syntax, such as ‘amusing critique of fixie culture, and . . . ’. Embedding the links in this way is a grammatical choice; the blogger could have highlighted the hyperlinks within the sentence using the deictic ‘here’, as shown in this sentence: ‘amusing critique of fixie culture, seen in the video here’. This would have extended the sentence, emphasizing the out-of-blog location of the video. Embedding the hyperlink within the ordinary syntax of the sentence, the blogger manages to almost inhabit the voice of the other text within his own. Bakhtin (1981) highlights how a writer can incorporate the voices of others in their own narrative through quotation or other forms of intertextuality; he refers to this as ‘double voicing’ or ‘ventriloquism’ and this seems to happen often on blogs where one text becomes embedded amongst others. The hyperlink enables this double-voicing to occur often online and when we see it as here, embedded unobtrusively, it is almost as if the writer is smoothing the edges of the texts together, blending the voices as much as possible. The hyperlinks are embedded in the sentence making them ‘tightly’ cohesive. As I discuss below, how bloggers exploit hyperlinks allows not only for textual cohesion and intertextuality, but also social networking.

Locality and Familiarity The first sentence of the blogpost: ‘Monday saw Bike Night at the Sheffield Showroom’ is a simple statement assuming familiarity with Sheffield as a place and that The Showroom is a cinema. Possible explanations why the blogger makes these assumptions include: he believes he is writing for a knowledgeable local audience; and/or he anticipates curious readers will explore his blog to seek understanding. Thus the blogger either assumes readers share the same local space, or he draws them to it, as they read further and discover not only where Sheffield is, but also see multiple posts about it, including photographs and video footage.

First person – seeing the world like me Early posts in this blog, which is just over a year old at the time of writing, include video footage taken from many other sites. However, recent posts share the experience of buying a camera to go on Gareth’s bike helmet, the difficulty of editing film, and then, examples of film that he has taken on his journeys around his locality. 235

HylandK_Chap15_Finals.indd 235

12/1/2010 8:07:05 AM

Bloomsbury Companion to Discourse Analysis

Figure 15.2 Video embedded in the blog (Gareth, 15 February 2010)

The post in Figure 15.2 shows the first frame of a film embedded in the post ‘Reasons to be fearful’. As mentioned, the title references the song ‘Reasons to be cheerful’ by substituting a word with similar sound but different meaning. This rhyming is a phonological cohesive device, and by drawing the words closer together Gareth simultaneously underlines the semantic difference. In turn, this emphasizes that he is about to show something that should NOT be fearful by inverting the meaning of the cheerful song he refers to. The language is thus subtly persuasive, drawing us into the writer’s cultural experiences of music and cycling where the semantics and Discourses of one impact on the other. Located above the written text, following convention, the reader assumes the film clip is intended to be watched first. The video lasts 75 seconds and is filmed using a small camera attached to the cyclist’s helmet. We thus share the cyclist’s view, and each bump in the road and environmental noise is witnessed via the soundtrack. Noticeably, there is no voiceover, nor any sense of editing 236

HylandK_Chap15_Finals.indd 236

12/1/2010 8:07:05 AM

Discourse and Computer-mediated Communication

having occurred; it is as if we are cycling in the blogger’s locality, experiencing with him, the dangers he is trying to illustrate. The camera angle provides a first person narrative; there is a sense of intimacy as we share aspects of the cyclist’s experience through sight and sound. This is consistent with the first person narrative of the written text throughout the blog, and also with the way we are drawn into the local space of the blogger, this time with his experience of the locality of Sheffield being embedded, as it were, within the virtual space. The written text is very specific about location: Sheffield Supertram tracks represent a real danger for cyclists. At any time there is the risk of a wheel dropping into one of the two parallel grooves, in the wet there is also the risk of sliding on the metal tracks. The best route to take is to cross the tracks at 90 degrees, but this is rarely possible. A particular problem is Hillsborough Corner which links two main routes out of the city to the North West. The video shows me negotiating the corner from South to North, but it is actually more difficult in the other direction as the road slopes uphill and turning from Langsett Road into Ripley Road to head up Walkley Lane is effectively an impossible manoeuvre to do safely when the road is wet. The details are precise and methodical; the lexicon is drawn from technical cyclist Discourse, for example: wheel, parallel grooves, metal tracks, 90 degrees, main routes, manoeuvre. We also see local knowledge coming to the fore with precise geographical naming of locations, such as: Hillsborough Corner, North West, Langsett Road, Ripley Road, Walkley Lane. Again, we are drawn into the locality as we are taken on this trip with the cyclist. As mentioned earlier, the representation of the local is not uncommon in online spaces and boyd has argued, The digital era has allowed us to cross space and time, engage with people in a far-off time zone as though they were just next door, do business with people around the world, and develop information systems that potentially network us all closer and closer every day. Yet, people don’t live in a global world – they are more concerned with the cultures in which they participate. (boyd 2006: 1) Nevertheless, the blogroll, the hyperlinks and the semantic content of a number of posts within this blog, display connectivity with bloggers in Canada, the United States and other parts of the United Kingdom – who also share information about their specific environments. There is a sense in which these bloggers seek to demonstrate through local concerns, an affinity with others and a connectivity with cyclists who are similarly discursively constituted within their own local worlds. There is a sense of dialogue across these bloggers when one 237

HylandK_Chap15_Finals.indd 237

12/1/2010 8:07:06 AM

Bloomsbury Companion to Discourse Analysis

reads about similar concerns about cycling round neighbourhoods across the world; an appreciation of the importance of the local within the global sphere. As Myers (2010: 77) explains, ‘Bloggers use a range of devices, direct and indirect, that can make readers feel like they are being talked to, included in a group, and involved in the blog.’ In this blog this is achieved, for example, through the direct address; the development of an in-joke around song-based post titles; the use of insider technical language; the use of ‘first person’ video narrative, sharing local knowledge and making it accessible.

Places to Link Myers identifies 3 main types of site to which blogs link: 1. Other blogs; 2. Mainstream media (e.g. BBC, the Guardian, New York Times) 3. Websites that are not blogs

Expanding on this I add: 2. Links within the same blog (self-referencing) 3. Links to other user-generated sites (e.g. YouTube, eBay, Flickr) I discuss each of these types of linking in turn.

Other blogs – displays of connectivity Digital technology facilitates the duplication and replication of all modes of text. Cross-fertilization of ideas online is rife and despite digital data and the ability to follow hyperlinks, it is often difficult to trace the origin of particular ideas, because of the complex weaving of multilayered texts. Ideas can spread via hyperlinks; through narratives that have meshed together; through images, all of which may be copied or altered in some purposeful or accidental way. In my observations, online individuals seek connectivity; they make rather than break chains, as we see in the example from blogger jozjozjoz below: Stolen from Skits, who stole it from Meeta, who stole it from Busy Mom, who stole it from Jill, who stole it from Neva Miss Feva, who stole it from 3rd House Party, who stole it from Twilight Cafe, who stole it from Pam, who stole it from Jo, who stole it from Whump, who got it from tamiam, and oh, I give up. Just do it. (Which I blatantly steal from Nike.) (jozjozjoz 2004) This extended sentence, which acknowledges the source of an idea jozjozjoz uses in his blog, is quite an extreme demonstration of connectivity. There is a definite ironic tone here, as he credits all those he can trace; he carefully shows 238

HylandK_Chap15_Finals.indd 238

12/1/2010 8:07:06 AM

Discourse and Computer-mediated Communication

how in attributing acknowledgement online, it can be a long complex process since so many texts are embedded within each other. The embeddedness of the subordinated clauses grammatically enacts the embedded nature of the idea as it has passed through the internet from person to person, text to text. The elaborate enactment, through the series of subordinated clauses, along with hypertext, is playful and emphasizes jozjozjoz’s online connectivity. Jozjozjoz also illustrates how it is possible to trace ideas online. Each underlined word on this text represents a hyperlink; one can link directly from this text to all those other bloggers this blogger has acknowledged. In this way the texts of those blogs become incorporated into jozjozjoz’s blog, and simultaneously they are drawn into a delineated affinity, where they are linked textually. Of interest also are the bloggers’ names, referred to here; in collecting these names together, not only does jozjozjoz provide a ‘public display of connection’ but also reflects the playfulness of many of the bloggers’ online nomenclatures. These range from the straightforward abbreviated names like Pam, or the abbreviated Jo, to onomatopoeic names like Whump, the descriptive Busy Mom, and the rhyming, creatively spelled Neva Miss Feva. All these contribute a ludic feel to this blogpost, but also betray an online identity of each respective blogger. We see how the blogging practices of many bloggers can contribute to one specific literacy event. Thus in outlining the provenance of his own work, jozjozjoz reflects on ways in which the narrative of his text links to other texts both structurally and semiotically. His final phrases, ‘Just do it. (Which I blatantly steal from Nike.)’ comprise an intertextual reference to an advertisement, and in making this reference, he demonstrates knowledge of popular culture and creates a sense of fun. Finally, by breaking the syntactic pattern, he emphasizes that this time he IS stealing, and not linking through a hyperlink back to the source of his citation. This is skilful writing since it shows control over how syntax works alongside meaning for deliberate effect, but which also shows understanding on another level, since there is a kind of deeper syntax, represented by the hyperlink, which is also co-opted into the semantic field.

Links to mainstream media and other sites The New York blogger and freelance journalist C-Monster, uses her blog as a way of communicating with potential employers – showcasing her writing skills as well as her ability to research. She focuses on the art world of New York. As well as other types of post, she regularly provides a Digest of art related events happening around New York as shown in this extract: The Digest. 03.22.10. ... □ An absolutely stunning/horrific photo essay and interview with the photographer from Cambodia’s notorious Tuol Sleng prison. 239

HylandK_Chap15_Finals.indd 239

12/1/2010 8:07:06 AM

Bloomsbury Companion to Discourse Analysis

□ Woot woot! The Is This Art? iPhone app is on Art Daily! Want to submit a one-liner crit for the app? You can do that right here.

□ Sorta related: Boxer shorts worthy of encyclopedic collecting by a venerable New York institution. (@TylerGreenDC.)

□ #class reviewed in the New York Times. Rawk on. □ For the Geeks: The @ symbol is now part of MoMA’s permanent architecture and design collection.

□ Artist Steve McQueen wants to put Britain’s Iraqi war dead on stamps. Images here. This post exemplifies Blood’s category, blog as a database of links. Her bullet pointed list has no introduction or summary; there is an assumption that readers should regard the blog like a catalogue to scan. The blog has an extremely long blog-roll and very regular densely written ‘Digests’ apparently designed for those ‘in the business’ who can use this as a ‘one-stop shop’. She demonstrates a cryptic ability to sum up, locate relevant information, and simultaneously express her opinion. For example, ‘#class reviewed in the New York Times. Rawk on.’ refers to an exhibition called ‘#class’ and links to a review in mainstream media, The New York Times allowing her to ‘double-voice’ this text; she closes with an invective exclamation ‘Rawk On’, possibly an imperative, using a common mis-spelling of ‘Rock’ (on), showing her enjoyment and sense of trend in the work. She refers to mainstream media at least half a dozen times in the post, as well as to major art galleries, but also to quirkier exhibitions ‘Boxer shorts worthy of encyclopedic collecting’. In this way, she showcases an eclectic mix, seeming like a curator of art for the expert. In just one post, C-Monster uses most of the roles identified by Myers (2010) that links can play in blogs; telling us more of what we already know – ‘An absolutely stunning/horrific photo essay and interview . . . ’; providing evidence for a claim – ‘The @ symbol is now part of MoMA’s permanent architecture and design collection’; giving credit – ‘(@TylerGreenDC.)’; leading to action – ‘You can do that right here.’, solving a puzzle posed by the lack of information in the linking text, ‘#class reviewed in the New York Times’.

Self-referencing links References within one’s own blog allows one to re-invoke old material; to establish the consistency of one’s performance of self; and to encourage close reading. Such links can also refer new readers to background information without needing to repeat it. For example, Guy in My Vedana writes, Well I finally got recognition of some of my multiple identities. How postmodern is that! So the doorplate saga rolls on . . .

240

HylandK_Chap15_Finals.indd 240

12/1/2010 8:07:06 AM

Discourse and Computer-mediated Communication

Here Guy refers to a previous entry (the doorplate saga) where he had described the discussion that had occurred around a new nameplate for his office-door after a promotion. The colloquial, ‘Well’, opening the post, seems to take up a continuing narrative; a discourse marker strongly evocative of spoken anecdotes and linking back to his previous post about the ‘doorplate saga’. His whole post, ironically looking at identity in the workplace, is couched in a complex of the colloquial alongside the academic, achieving an ironic tale that readers can trace back through his hyperlinked references to previous door photographs.

Blogs as Hubs of Digital Identity Someone who regularly frequents many sites can bring together the ‘digital threads’ of spaces to their blog. For example, my own blog acts as a kind of digital hub: As well as providing a space for me to discuss my academic work, the blog shows my latest contribution to Twitter (‘Finishing my chapter’ (!)); latest

Figure 15.3 The Digital Identity Hub

241

HylandK_Chap15_Finals.indd 241

12/1/2010 8:07:06 AM

Bloomsbury Companion to Discourse Analysis

images on my Flickr stream; and a hyperlinked ‘badge’ linking to an online research group. As well as making it easy for me to navigate to these sites, these links (or widgets) provide additional research-based identity markers; they are multimodal semiotic resources enriching the research Discourses of the blog.

Conclusions and Future Directions Discourse analysis of CMC involves looking at patterns within and across texts; I have argued that in order to take full account of ways in which communication takes place online, we need to focus not just on linguistic features, but also upon multimodal aspects and the wider sociocultural context. This wider sociocultural space may be the geographical locality of the text producers, or may refer to other online spaces. It is often the case that interactants develop jokes or ways of being together and this can provide a sense of a new online locality. This chapter has looked at the ways in which semiotic resources are drawn together for realizing certain types of performance within blogs. Virtual spaces are new contexts for interaction and research; the opportunities these contexts offer to linguistic ethnographers are rich and varied. Future work might look for example at ways in which individuals use profiles in social networking templates – such as in blogs, Flickr.com, eBay or LinkedIn. Discussion forums are spaces where alongside conventions being prescribed (set out by moderators), users also develop their own conventions. Analysis of the prescribed and non-prescribed ways of interacting in forums is another area for analysis. Politeness strategies, introduction rituals and ways of disagreeing, have particular protocols in offline spaces, which seem slightly different online and which come to the fore in discussion forums; these areas are again likely to be rich research contexts. Virtual Worlds (VWs) (like Second Life) are attracting researchers from many disciplines; how people communicate multimodally in VWs is likely to be of interest, and areas of similarity and variance to offline spaces are again likely to provide rich data. Spaces like Twitter, which allow only very concise messages (of 140 characters) invite questions around the strategies people use to be concise, and how they enrich their text with links to other texts.

Notes 1. eBay is an online shopping site. 2. Flickr is a photo-sharing website. 3. A feed brings to one site, all new posts from blogs to which someone subscribes.

242

HylandK_Chap15_Finals.indd 242

12/1/2010 8:07:07 AM

Discourse and Computer-mediated Communication

Key Readings Bruns, A. and Jacobs, J. (eds) (2006), Uses of Blogs. New York: Peter Lang. Davies, J. (2008), ‘Pay and display: the digital literacy practices of online shoppers’, in M. Knobel and C. Lankshear (eds), Digital Literacies: Concepts, Policies and Practices. New York: Peter Lang, pp. 227–48. Kress, G. (2010), Multimodality: A Social Semiotic Approach to Contemporary Communication. London: Routledge. Myers, G. (2010), The Discourse of Blogs and Wikis. London: Continuum.

243

HylandK_Chap15_Finals.indd 243

12/1/2010 8:07:07 AM

16

Forensic Discourse Analysis: A Work in Progress John Olsson

Chapter Overview Origins of Forensic Discourse Analysis The Focus of Forensic Discourse Analysis The Location of Language within Sociolinguistic Interaction Observations on the Structuring of Forensic Discourse The Construction of Admission Textual Provenance and Mode Western Australian Royal Commission into Police Corruption Contemporary Interrogation Methods Consensus in Court Summary Notes Key Readings

245 246 247 249 250 253 254 255 257 259 259 259

If you have just arrived at this page having read other chapters in this volume, it is this author’s suspicion that what you are about to read may come as something of a surprise to you, given that it makes no theoretical statement, no claims for any branch of applied linguistics, and few concrete pronouncements as to any future directions of the area known loosely as ‘forensic discourse analysis’. The reason for this is simple: forensic discourse analysis is a rapidly expanding field with few visible borders. It is – to coin a phrase – a work in progress. The parent term ‘discourse analysis’ is now applied to a broad range of linguistic pursuits, from the mapping of syntactic structures to informational 244

HylandK_Chap16_Finals.indd 244

12/1/2010 8:07:27 AM

Forensic Discourse Analysis

content; from pragmatics to functional grammar to conversation analysis (CA), and from ethnography to rhetoric – to name a few. Outside of linguistics, discourse analysis has attracted interest from different branches of psychology to anthropology, to studies of subjective logic. As a result, ‘discourse analysis’ is now attached to a wide variety of prefixes, including ‘institutional’, ‘forensic’, ‘critical’ and so on (Coulthard and Johnson 2007: 7). So far there have been only attachments and few indications of any impending divorces. Just as in the wider domain of linguistics (and elsewhere) many scholars are interested in ‘discourse analysis’, so too within forensic linguistics a number of practitioners can be said to be pursuing branches of what is loosely termed ‘forensic discourse analysis’. Given the breadth of use of both terms ‘discourse analysis’ and ‘forensic discourse analysis’, the task of summarizing the latter field, both as an expression of the former and as a sub-domain within forensic linguistics, is a charge filled with the twin pitfalls of generalization and exclusion: the danger, on the one hand, of generalization of the field to the point of not being able to give specific information, and, on the other, the necessary exclusion of highly specialized, though equally worthwhile, fields of endeavour through a lack of space. Summarizing current thinking and research on these topic areas is an even more daunting task, not only because of the somewhat heterogeneous breadth of work associated with the field, but also because many of those working in areas which may be candidates for the umbrella term of ‘forensic discourse analysis’ might not use – or even be prepared to recognize – this term to describe what they are doing.

Origins of Forensic Discourse Analysis Notwithstanding these caveats, the origins of ‘forensic discourse analysis’ can be traced to a chapter in a book by Coulthard, in which he appears to describe the subject as the analysis of spoken discourse in forensic settings, with particular reference to police interviews (Coulthard 1992: 242–3). In a 1994 paper on the Derek Bentley statement, he refers to the use of insights from a number of linguistic disciplines collectively as ‘forensic discourse analysis’ (hereafter FDA), including the use of data from prescriptive grammar (word order of particular phrases, especially the postposed position of ‘then’ in police statements), corpus linguistics (word frequencies and collocations) and functional grammar (especially issues of register, specifically language mode (Coulthard 1994). However, since that time the field has expanded considerably and the term is now applied with a catholicity not perhaps anticipated by its early practitioners. In the light of this it may be time to review some areas of progress. 245

HylandK_Chap16_Finals.indd 245

12/1/2010 8:07:28 AM

Bloomsbury Companion to Discourse Analysis

The Focus of Forensic Discourse Analysis The main focus of the analysis of forensic discourse in the early days was in the area of police interaction with suspects and defendants. This will also form the bulk of this chapter, though not exclusively. Other areas which have attracted linguists’ attention include court examination and cross-examination, language use in lawyer–client relationships and the analysis of a variety of text types. A number of researchers, from traditional discourse analysts to conversation analysts and sociolinguists, have been engaged in these endeavours. Coulthard’s 1994 paper set the benchmark for early textual analysis of written police statements. In that paper, Coulthard gives cogent reasons for the suspect status of the word ‘then’ in the Bentley statement. He begins by pointing out that in statements known to have been authored by police officers ‘then’ occurs approximately once per 100 words, whereas in witness authored statements its frequency is about 1/10 of that. This is borne out by his examination of the Cobuild corpus which gives a 1/500 word frequency. However, it is not just the frequency of ‘then’ which appears curious, but its positioning: in the Bentley statement it is usually placed after the subject, rather than before, which is more usual in ordinary speech, so that instead of getting ‘then I . . . ’ we get, in the Bentley statement, ‘I then . . . ’ (Coulthard 1994: 32). Coulthard noted that this phenomenon also occurred routinely in police-authored statements. While still working on the Bentley statement, in a later publication (Olsson 2000) I collected ten dictated narratives from ordinary speakers in order to evaluate the claim that the statement had been dictated. As part of the analysis, issues of textual cohesion, lexical distribution and proportionality were considered. It was found that the Bentley text exhibited odd characteristics in these respects. Regarding cohesion it was noticed, for example, that the name ‘Chris’ is used ten times in the statement, but that the subject pronoun ‘he’ occurs only four times in relation to ‘Chris’. On reading the text, the use of ‘Chris’ seems somewhat overdone, if not artificial. In any case, once a participant is within a given universe of discourse it is not necessary to continue naming them. It is a general characteristic of police-authored statements that they tend to specify individuals by name, and – in general – do not use the available resources of textual cohesion, such as the pronoun system. It is also curious that this same participant, ‘Chris’, is referred to as ‘Craig’ in the first part of the text and only becomes ‘Chris’ in the second and third parts of the text. Proportionality in text construction relates directly to frequency. The most frequent lemma in the text is ‘policeman’ (ten times), with its variant ‘policemen’ occurring once. Even though this lemma (roughly the dictionary headword and any variants thereof) is the most frequent item in the text it 246

HylandK_Chap16_Finals.indd 246

12/1/2010 8:07:28 AM

Forensic Discourse Analysis

makes its first appearance at well past the halfway mark in the text. In relation to most texts, this distribution is singular.

The Location of Language within Sociolinguistic Interaction FDA practitioners include those with an interest in, inter alia, sociolinguistic interaction. For scholars like Eades, the emphasis is on situating interaction within its social context. After studying the interaction and communication styles of Aboriginal Australians for some years (Eades 1988), Eades came to realize that in institutional settings Aboriginals would be at a severe disadvantage because of fundamental differences between European and indigenous Australian ways of asking and answering questions. In Aboriginal culture, agreement and consensus are formed as a group endeavour, in gradual stages. There is an avoidance of conflict and a desire for social harmony. The confrontational style of direct questions and demands for ‘straight’ answers favoured by non-Aboriginals is entirely absent from their culture. Thus, when Europeans and Aboriginals meet in institutional contexts there are inevitably clashes. Eades illustrates this with examples from the well-known Pinkenba case: In May 1994, three boys were picked up by police and driven to a remote location where they were abandoned. They had apparently not committed any offence, and nor were they ever charged. However, charges were eventually brought against the officers for depriving the children of their liberty. The main issue at the preliminary hearing was the question of consent. While police maintained that the boys had consented to get into the car, prosecution claimed they had been coerced. At Pinkenba the boys stated they had been forced to strip naked and had been threatened with being thrown into a creek. Eades’ analysis of the courtroom interaction reveals the one-sided nature of the proceedings. Not only are the boys – the victims in this case – frequently referred to as the ‘defendants’ – but they are ceaselessly badgered by the defence counsel. The case did not go further than the preliminary hearing because the magistrate found that the boys had ‘consented’ to go in the police cars. The court not only did not take into account that the lawyers in the case were able to bully the children in the witness box, but it also failed to appreciate the fact that in Aboriginal society young people are obliged to respect their elders, and that part of this respect involves avoidance of interactional conflict. The key point to emerge from Eades’ study is that in what are essentially ‘white’ courtrooms the neocolonial interaction style is dominant. Those who fail to understand it or use it are at an inherent disadvantage (Eades 1995: 10). The fact that many courts do not recognize that there are differences in interactional style between the colonizers and the colonized means that the existence of any disadvantage is not only denied, but officially sanctioned. 247

HylandK_Chap16_Finals.indd 247

12/1/2010 8:07:28 AM

Bloomsbury Companion to Discourse Analysis

At a linguistic level, the failure to recognize differences between Aboriginal English (AE) and Standard English (SE) also extends, not surprisingly, to statements allegedly dictated by indigenous suspects to European police officers. Eades (1995) relates the cases of Rupert Max Stuart and Kevin Condren, both convicted, in Australian courts, of murder. In the first case, Stuart was convicted in 1959 of the rape and murder of a 9-year-old European girl, Mary Hattam, in South Australia. Initially sentenced to hang, this was later commuted to life imprisonment. In the second case, Condren was convicted of the 1983 murder of Patricia Carlton in north Queensland. He was also sentenced to life imprisonment, of which he served seven years. Eades did not herself undertake an analysis of Stuart’s confession. This had already been carried out by T. G. H. Strehlow, the well-known linguist and anthropologist who had been raised on a mission station in the same area as Stuart, and who was a fluent speaker of the Aranda Aboriginal language. As Eades points out, Strehlow carried out a forensic linguistic analysis, at least ten years before anybody in Europe or America did. Moreover, he took a discourse approach long before the concept was even clearly understood in linguistics. Strehlow’s advice had been unambiguous: Stuart’s confession could not have been spoken by a speaker of AE (or what was then termed Northern Territories English, NTE). He gave two simple indicators. The use of unsupported verbal auxiliaries (not found in other examples of the defendant’s language or in AE or NTE generally): The police statement contained formulations such as: ‘When you hit her with the steel picket, did you aim for her head?’ ‘Yes, I did.’ ‘Was Patricia bleeding when you walked away?’ ‘Yes, she was.’ (Eades 1995: 154) The use of the definite article, the: Whereas the police statement has examples like ‘Alan went back to the showground’ and ‘I slept at the Wheatboard’, Strehlow recorded examples such as ‘Alan went back to showground’ and ‘I slept in that Wheatboard’. As a result of these and other observations of basic grammatical differences, Strehlow concluded that the grammatical features of Stuart’s language were inconsistent with the language of the police interview. One particularly absurd invention attributed to Stuart was: ‘The show was situated at the Ceduna Oval’ whereas the Aboriginal speaker himself stated in his conversation with Strehlow ‘That show was in Cedoona Opal’. 248

HylandK_Chap16_Finals.indd 248

12/1/2010 8:07:29 AM

Forensic Discourse Analysis

Observations on the Structuring of Forensic Discourse However, Strehlow moved beyond mere grammatical analysis. He also concluded that it would have been impossible for someone not educated in the construction of English prose to produce a ‘narrative [with such] conciseness and relevance’ (Inglis 1961, cited in Eades 1995: 158). Strehlow’s concern was not just with the ‘discourse features of the “confession”, but with the “whole composition” ’, which he claimed had the ‘structure and argument of a legal document’ (Eades 1995: 158). Although nearly 30 years separated Strehlow’s analysis of Stuart’s supposed confession and Eades’ examination of Condren’s ‘confession’, Eades found a number of similarities between the communication styles of the two alleged confessions. For example, both confessions were highly focused on specifying times, dates and other numeric information whereas, in reality, neither Stuart nor Condren appeared capable of thinking in terms such as these. So, when asked what time he (allegedly) attacked his victim, Condren is supposed to have answered, promptly: ‘Quarter past four’. However, when asked in prison what day a particular incident had occurred Condren is not able to be any more specific than ‘Um, about, about two weeks ago’. In court, his incompetence with numbers is further demonstrated. Asked how many flagons of wine he had drunk, he replies ‘About twelve, I think, or seven’. Strehlow had also noticed that Stuart’s confession was full of references to time, even though Stuart was almost entirely uneducated (Eades 1995: 161). These observations are in line with what has been noted extensively by forensic linguists elsewhere: namely, that police officers appear to be devoted to the precise recording of time and other quasi-numerical data. For example, Fox (1993) writes: Police officers are obsessed by time. Or so it seems from their statements. Actual times are often given: ‘at 5.12 p.m.’, ‘at 9.23 p.m.’, ‘at 12.46 p.m.’ etc. These are frequently the times at which questioning begins and ends, but by no means only that; for example, ‘at 12.20 p.m., at a rubbish site . . . ’; ‘Lunch was provided at 12.39 p.m.’ There are also many approximate times: ‘at approximately 3.15 p.m.’, ‘at 10.28 a.m. approximately’, ‘at about 3.45 p.m.’, ‘at round about 10.05 a.m.’ (Fox 1993: 183) Although some might question whether commentary on references to time can be classified as a truly discourse analytic topic, its importance should not be underestimated. The Western preoccupation with time has largely come about because of the modern individual’s growing involvement with official institutions, particularly arms of administration, such as government offices, courts and other official bodies. This Western adherence to matters of time 249

HylandK_Chap16_Finals.indd 249

12/1/2010 8:07:29 AM

Bloomsbury Companion to Discourse Analysis

has caused a general awareness of the day being ‘divided up’ into distinctive segments relating to the kinds of activity people carry out in order to function as productive citizens. In the West, events are always related to their time of occurrence, and this is ingrained within our discourse. Amongst indigenous Australians, however, events are related not to time but to other events, especially those which relate to families. People remember events not in relation to when, but who. Therefore, insignificant as it may appear . . . as it may appear in the construction of police statements, the question of time is a key dividing line between the cultures of those who measure their lives in days, hours, minutes and seconds, and those whose lives are built around their family relationships and the events that take place within those families.

The Construction of Admission It was, therefore, with some surprise that I recently reviewed the police statements in another famous Australian miscarriage of justice, namely the Mickelberg case. In 1982 three brothers, Raymond, Brian and Peter Mickelberg were arrested on suspicion of having carried out the infamous Perth Mint swindle. It was later found that the lead detective had persuaded a colleague to fabricate statements in which the brothers made admissions, which led to lengthy prison sentences – later overturned on appeal. In that particular case I analysed approximately 50 statements, yet I found very few statements which contained a mention of time. Given the general police obsession with time – whether in the United Kingdom, Australia, the United States or elsewhere – this was a striking feature of the Mickelberg statements. What was also striking in the Mickelberg case was that there are many examples of unforced self-incrimination in the statements. The typical scenario is that the defendant initially denies an action or situation, but then – on further questioning – admits the action or situation, either in its entirety, or partially. Here is an excerpt from an interview between Don Hancock (a senior police officer, now deceased) and the suspect Raymond Mickelberg, accused of perpetrating the Perth Mint fraud:

Example from the Mickelberg Case Statements I (Hancock) said, ‘Does Peter know that you are Peter Gulley?’ The accused said, ‘Look, Peter doesn’t know anything. I swear to you, none of the family are involved.’ I said, ‘What about Brian?’ 250

HylandK_Chap16_Finals.indd 250

12/1/2010 8:07:29 AM

Forensic Discourse Analysis

The accused said, ‘I told you, the family have got nothing to do with it.’ I said, ‘Are you saying that you did it all yourself?’ The accused said, ‘Look, I’d like to tell you, but I can’t say anything.’ (Hancock interview of Ray Mickelberg 15 July 1982) One of the key discourse strategies for moving from a position of claimed denial by the suspect to apparent concession or admission is the use of vagueness. Thus, although the question asks whether Peter knows anything about Gulley, it is not stated in the answer what Peter does not know anything about. Rather, the answer simply states ‘Peter doesn’t know anything’. This is then broadened from Peter to the entire family and from ‘knowing anything’ to having ‘nothing to do with it’. Whatever ‘it’ is has now become the focus – the attention having completely shifted from Peter Gulley and what Peter knows. The next question takes up ‘it’ in the form of ‘ . . . that you did it all by yourself’. Note that this is an embedded question. The answer ‘I’d like to tell you . . . ’ implicitly accepts the proposition that the speaker ‘did it’, whatever ‘it’ – unspecified – is. The accused now repeats a standard mantra, found in many of the statements that he ‘can’t say anything’. We thus have a number of non-content pronouns: it, anything, nothing: in fact, nothing is being specified. However, the reader is deliberately left with the impression that someone other than the suspect is involved in ‘it’ because he ‘can’t say anything’ about whether others are involved in ‘it’ or not. All of this has stemmed from the question as to whether Peter knew that Raymond was Peter Gulley. I suggest that Peter Mickelberg knowing who Peter Gulley was is of no relevance to the Perth Mint fraud inquiry: it is simply a red herring. By using vagueness, the writer of the statement is able to suggest that the witness has begun by deliberately attempting to conceal information from the officer, but has almost immediately succumbed to the officer’s close questioning. Such an exchange, read out to a jury, can be very damaging to a defendant.

Indications of Fabrication However, close examination of the statement reveals a strong possibility that it is fabricated. According to the detective Hancock’s account, he conducted the interview with Raymond Mickelberg, while his colleague, Sergeant Round, noted down the interview (we will discuss the supposed method by which this was done shortly). Later, when he had to prepare his own statement, Hancock admits that he copied his version from Round (altering only such items as ‘Sergeant Hancock’ to ‘I’, for example). This method of interviewing 251

HylandK_Chap16_Finals.indd 251

12/1/2010 8:07:29 AM

Bloomsbury Companion to Discourse Analysis

and statement making was not unusual in Western Australia at the time. However, close analysis of the two statements reveals, not that Hancock copied from Round (as the officers claimed), but that Round most likely copied from Hancock. Hancock’s statement reads, with reference to a building society passbook: Hancock version: He said, ‘I thought it was that long ago.’ I said, ‘Do you remember operating the account at Whitfords on the 27th of May?’ He said, ‘Yes.’ I said, ‘Why did you get a cheque made payable to Mr Wilson?’ Round version: The accused said, ‘I thought it was that long ago.’ The accused said, ‘Yes.’ Sergeant Hancock said, ‘Why did you get a cheque made payable to Mr Wilson?’ It appears that in copying out the Round version the writer has omitted the line ‘Do you remember operating the account at Whitfords on the 27th of May?’ thus producing the anomaly of the suspect apparently answering a question without a question being asked (note two lines in succession begin ‘The accused said’). Thus, it does not seem likely that Hancock would have inserted the sentence ‘Do you remember . . . ’, but that Round accidentally omitted it when copying. This is supported by the observation that a reference to remembering the cheque cashed at Whitfords PBS exists in Round’s own notes of the alleged Hancock interview. Yet it is missing from his statement. This also raises the question of Round’s own statement and which version of notes he copied from when producing it. It seems possible that he wrote out his own rough notes only after he had written his own statement. Another difference relates to a line found in Hancock’s statement and also in Round’s: Hancock: ‘ . . . why would I use my own number on the cheques?’ Round: ‘ . . . why would I use my own number on the cheque?’ This point relates to a very important facet of the case and a key element of the prosecution. The issue of whether it is ‘cheque’ or ‘cheques’ is central to the inquiry. As there were several cheques used in the offence, then ‘cheques’ would appear to be the correct form with ‘cheque’ being the incorrect form. 252

HylandK_Chap16_Finals.indd 252

12/1/2010 8:07:29 AM

Forensic Discourse Analysis

Round himself is aware that there is more than one cheque, because he says in his statement: Round statement I said, ‘Two of the cheques used to buy the gold were issued by the Perth Building Society and were drawn on the account of Peter Gulley.’ Furthermore, in his own notes of the alleged Hancock interview he writes ‘cheques’: ‘ . . . why use own number on cheques?’ Again, this would point to Round not using his own rough notes to compile his statement, and from that for Hancock to produce his statement. Rather, it suggests he has used some other version when writing his statement. I suggest that – given the suspect identical sentence across the two alleged interviews – the likely source of the ‘cheque/s’ error is the copying procedure from another set of notes or statement other than his own rough notes. If Round had originated his own document, he would most likely have been aware of the ‘cheques’ versus ‘cheque’ issue and would have written ‘cheques’. However, when copying it is relatively easy, especially in a long or tedious document, to make small errors. Hancock also knew that there was more than one cheque, and in fact writes ‘cheques’: if he had been copying from Round he would most likely have pointed this error out to Round, since it is a very important detail in the case that several cheques were used rather than one.

Textual Provenance and Mode An issue which has come up repeatedly in forensic discourse analysis is that of provenance and mode. While police officers claimed that statements such as those which featured in the Bentley case, or in the Birmingham Six and Guildford Four cases had been dictated, some Australian police officers of the 1970s and 1980s appear to have taken a more creative approach to the art of explaining the source of ‘confession’ statements. The first claim to consider, but one which would hardly merit investigation in the real world, is that which applies to the statements in the Mickelberg case. Here the officers claim that they took the suspects’ confessions and admissions down ‘verbatim’ in long hand. We can dismiss this almost immediately by noting Hargie and Dickson’s observation that whereas people ordinarily speak at a rate of somewhere between 125 and 175 words per minute (Hargie and Dickson 2004: 184), adult handwriting speeds generally average between 22 and 31 words per minute. It is therefore extremely unlikely that the average adult can take down conversation verbatim for any length of time. 253

HylandK_Chap16_Finals.indd 253

12/1/2010 8:07:29 AM

Bloomsbury Companion to Discourse Analysis

The officers in the Mickelberg case appear to have attempted to overcome this difficulty by proposing an entirely novel way of text construction, which I have provisionally termed ‘sequential transcription and turn’. According to the officers’ testimony at the trial the method works like this: The officer conducting the interview asks a question. The suspect does not answer it until the scribe (another officer) has written down the question. Only once the question has been written down does the suspect answer the question, which the scribe in turn writes down. If the scribe cannot keep up with the conversation s/he asks the interactants to repeat or to slow down. I consider this method entirely implausible and linguistically naïve. It not only reinforces the much discredited, speaker-hearer model of the 1960s, but it also proposes, in effect, that speech is not interactive, but sequential, which it rarely is. In general, people do not wait until someone finishes speaking before answering. Interruptions are not only commonplace in conversation – they are almost inevitable at one or more stages in the conversation. Secondly, one aspect of police interviewing is to test the suspect/witness’s reaction to a question or the mentioning of some event or person involved in the inquiry. Using the method of sequential transcription the officer cannot ‘surprise’ the suspect with a question. The suspect would have to wait until the scribe officer had finished writing before answering. This would be a severe disadvantage to the police officer because the suspect would have considerable time to formulate an answer. When we ask someone a question, the likelihood is that they will understand the question before we have finished asking it. In many cases they will want to answer it even before we have finished asking it. Humans process speech at a very high rate. It is not natural for speakers to wait 20 or 30 seconds before they can answer a question. One implication of this model is that the officer prepares the next question while the suspect is answering and the scribe is writing down the answer. In other words, the implicit claim is that the entire interrogation is planned beforehand. However, following questions are often based on current answers and so the officer’s ability to plan is much less than the model suggests. I suggest the method of sequential transcription proposed here would be as linguistically disorienting to officers as it would to interviewees. All in all this method of interviewing has little linguistic credibility and I suspect few linguists would believe that any of the statements were produced in this way.

Western Australian Royal Commission into Police Corruption In 2002, a Royal Commission was formed to investigate corruption in Western Australia’s police service. At this commission police officers, under the 254

HylandK_Chap16_Finals.indd 254

12/1/2010 8:07:30 AM

Forensic Discourse Analysis

protection of anonymity, testified that they had verballed suspects on many occasions and, also, that they had perjured themselves on many occasions. One officer admitted to having altered evidence in the Mickelberg case. In another hearing, Sergeant Anthony Lewandowski admitted that he and Detective Sergeant Don Hancock had assaulted Peter Mickelberg. At the Royal Commission hearings several officers described how they would make their verbals appear more realistic. One officer would engage the suspect in friendly conversation in order to acquire an understanding of their speech habits, mannerisms, hobbies, topics of interest and information about their families. This information would then be used to construct a statement, that is to verbal the suspect. In some parts of the Western Australian Police, according to several of those who testified, promotion was sometimes dependent on the officer’s willingness to verbal, and – on occasion – to assault suspects. In an affidavit regarding his part in the Mickelberg case, Lewandowski stated, inter alia, ‘The statements we made up were based on later information . . . ’ and ‘Don [Hancock] and I just sat around adding in what we reckoned we needed. . . . we thought we’d lose Brian [Mickelberg] because we didn’t tell enough lies . . . all the . . . evidence in relation to the so-called confessions of Peter . . . Raymond . . . and Brian Mickelberg . . . was false. The statements were fabricated by Don Hancock and myself sometime in early September 1982’ (WA Police Royal Commission, 2002: Section 3.8 The Practice of Verballing).

Contemporary Interrogation Methods There is a general perception among lawyers that since the advent of audio and video recording of police interviews, linguists and other professionals need not trouble themselves any longer with issues of verballing or other forms of evidence tampering, including oppressive interviewing. However, despite the undoubted benefits of audio and video recording to the interview process there still remain three areas of concern: Transcription: Not all police interviews are transcribed in full – many are summarized by police typists who have little or no training in linguistics. It is rare that an interview tape would be played in court, and it is thus very often the case that the only record of interview presented to juries is an interview summary. In my experience, solicitors very seldom take the trouble to check these summaries, or even to discuss them in any depth with their clients. However, legal professionals need to be mindful that even though everything may appear to be above board, the document that is being produced is still an institutional police version of events and may be wittingly or unwittingly 255

HylandK_Chap16_Finals.indd 255

12/1/2010 8:07:30 AM

Bloomsbury Companion to Discourse Analysis

slanted against defendants. In a recent case the officer conducting the questioning asks the suspect how long he had been in possession of a particular vehicle. The suspect replies: ‘Not long’. In the summary this is rendered as ‘X denied that he had possessed the vehicle for a long time’. Shortly thereafter the suspect is asked if he was at a certain address on a particular date. He says ‘No, I wasn’t. I was at . . . ’ and names another place and an alibi. The alibi is not mentioned in the summary. All that is stated is ‘X denied that he had been at . . . on the date in question’. The use of denied in this way can be particularly injurious to a defendant in court. If the summary were read out in court it might lead a jury . . . that the witness was lying, being uncooperative or simply being evasive.1 It is, therefore, very important that solicitors check summaries against interview tapes in order to ascertain that a fair summary has been made, and that the defendant’s words have not been (un)wittingly misrepresented. Memory construction/oppressive interviewing: In an ongoing appeal case in the Australian courts a man convicted of murder claims that police placed photographs of the scene of the crime and the weapon in his full view in the interview room, but out of sight of the recording cameras. Police, naturally, deny this. However, of interest is that in the course of the interview the interrogating officer asks the suspect, on several occasions, to close his eyes and ‘try to remember that night’, ‘remember . . . what happened in that room’, ‘take your time’ and ‘remember . . . what you did in that room’. Whether the appellant’s challenge to the conviction will ultimately be successful or not, it is clear from the interview record that the above instructions, namely for the suspect to close his eyes and remember the night in question, border on an invitation to construct memory. Discursive paths to consent in police interviews: In cases where the actual language spoken in a police interview is not at issue because a recording is available, CA (see Wilkinson and Kitzinger this volume) has proved useful in demonstrating the nature of the interaction. Heydon (2005) has contributed much in this area. In one interview Heydon notes that the officer is able to produce agreement to a proposition. The officer begins by asking the suspect whether there was any reason for the assault he has (allegedly) carried out on his girlfriend: Officer: Was there any reason why you had to treat her this way at all?2 Suspect: It was a combination of things you know. I didn’t like the fact you know here I am going out with a girl and she jumps into bed with one of my so-called mates. As Heydon points out the officer implies by the use of ‘any reason . . . at all’ that no reasonable explanation for such behaviour exists. Hence, whatever explana256

HylandK_Chap16_Finals.indd 256

12/1/2010 8:07:30 AM

Forensic Discourse Analysis

tion the suspect may produce has already been signalled to be inadequate. This is borne out by the officer’s reply to the above, where he says: Officer:

OK well regardless of that may have been the case . . . not to say whether that might have been the case, do you agree that that warranted your actions by dragging her out by the arm, pulling her by the hair, forcibly removing her from the house?

Suspect:

No. (Heydon 2005: 142)

Excerpt from a Police Interview The turn beginning ‘ . . . regardless of that may have been the case . . . ’ functions, as Heydon notes, to ‘deny the relevance of the suspect’s contribution’ and, further, the officer then systematically ‘discards [the suspect’s] reasons why he assaulted . . . ’ the victim. In this way, the officer produces a version of events which disregards the suspect’s version. Although the suspect is able, in a limited way to initiate topics, the officer is able to ignore them, and to produce the final ‘favoured’ version (Heydon 2005: 142).

Consensus in Court It is not just in police interviews that legal institutions are able to manufacture consent. Several linguists have been working on courtroom discourse, in particular the discursive methods employed by barristers and attorneys in directing questions and controlling the information flow which is presented to the court. Lowndes (2007) presents a summary of some lawyerly methods. One in particular is that of propositional loading or stacking. This is the technique whereby a lawyer will present a series of questions and then ask for agreement at the end of them, often by use of a pseudo-tag question. The witness or defendant does not know what s/he is being asked to agree to. In this example from Lowndes, a barrister is cross-examining a medical expert witness regarding the injuries being claimed in a medical insurance case: Barrister:

and he’s failed them because when asked formally to show a range of movements, he showed a restricted range but he’s been observed informally or in a different context he reveals that he can actually do all the things he says he can’t do. That’s what the tests show, isn’t that right Doctor?

As Lowndes notes: ‘Pseudo-tags can follow single or multi-propositional turns and it proved difficult for witnesses to gauge whether [the word] that was a 257

HylandK_Chap16_Finals.indd 257

12/1/2010 8:07:30 AM

Bloomsbury Companion to Discourse Analysis

substitute for the entire original question or referred to the final proposition only’ (Lowndes 2007: 308). In rape and sexual assault trials other strategies are used. As Aldridge and Luchjenbroers (2007) note, a frequently used lawyerly ploy is to portray the alleged rape victim as someone with previous sexual experience, while at the same time minimizing any suggestion of coercion or violence. In one case a 14-year-old girl has allegedly been assaulted by a male wielding a knife. This is the exchange between the defence barrister and the child: Barrister: Witness: Barrister: Witness: Barrister:

Have you ever done anything like that with anyone else before? No. Has anyone ever fingered you before? No. So, that was your first sexual encounter was it?

Note the phrasing of the first question: ‘Have you ever done . . . ?’ The lexical choice of ‘do’ here is significant. Since the witness is purporting to be a victim, how can the use of ‘done’ possibly be justified? If the witness was a genuine victim, then the witness was ‘done to’, and did not ‘do’. As Aldridge and Luchjenbroers point out, the use of the word ‘fingering’ is also significant. It implies that the witness has a familiarity with sexual practices. The final question ‘So that was your first sexual encounter was it?’ minimizes the violence of the act with a knife. In the real world it is irrelevant whether this is the child’s first sexual experience. However, by phrasing the question like this the barrister disempowers the witness as a victim and portrays her as a sexually experienced individual who is an agent in the ‘encounter’ because she ‘did’ (‘have you ever done . . . ?’ (Aldridge and Luchjenbroers 2007: 93). The twin strategies of degrading the victim to the status of a willing, skilled participant and reducing the rape or assault act, with all its attendant violence, to a consensual experience are common lawyerly tactics in rape and child sexual assault cases. The authors point out that, despite the many initiatives in the United Kingdom and elsewhere to protect vulnerable witnesses – especially rape and sexual assault victims (both adult and child) – lawyers are still able to exploit language structures and lexical choices in order to effect their ends. In rape cases the social frame dictates that the victim is ‘respectable’ and that, ideally, the perpetrator is a stranger (Clark 1992, quoted in Aldridge and Luchjenbroers 2007: 90), whereas in child sexual assault cases, the child must be ‘innocent’ and entirely ‘inexperienced’. There is no penalty to the defence barrister for suggesting otherwise. Furthermore, rape victims must demonstrate not only that they are sexually respectable, but also that they resisted to the utmost the assault of their attackers. Part of the lawyer’s strategy is to represent anything less than the utmost 258

HylandK_Chap16_Finals.indd 258

12/1/2010 8:07:30 AM

Forensic Discourse Analysis

physical resistance as consent (despite the fact that this may put the victim in severe danger), which is achieved through linguistic means and a reliance on stereotypical frames (Aldridge and Luchjenbroers 2007: 104).

Summary The work reported in this chapter represents only a small sample of the kinds of areas in which linguists have been applying their knowledge of discourse to the forensic arena. Many other aspects of legal language have come under the microscope, including the wording of laws and statutes, the directions given by judges to juries, the nature of lawyer-client discourse in pre-trial discussions, the reporting of sexual crimes in the news media, the use of interpreters in asylum cases, to name but a few. I consider that the field of forensic discourse analysis is becoming wellserved by a growing body of researchers who are applying fundamental linguistic knowledge and skills in order to bring all areas of language and the law under the microscope. This is evidently a healthy field of linguistic endeavour. At the same time more lawyers and judges are beginning to realize the potential value to the legal profession of linguistic expertise. Reason dictates the need for this kind of cross-discipline cooperation to become more extensive, in order to reduce abuses and to ensure that the legal systems of all countries can begin to live up to their frequent claims of due process, impartiality and universal fairness. This is the true task of forensic discourse analysis: in that respect, consider the current chapter to be an interim report.

Notes 1. For an excellent introduction into interview resistance strategies, see Newbury and Johnson (2006). 2. See Heydon 2005: 142 for the original CA mark-up of this text.

Key Readings Cotterill, J. (2003), Language and Power in Court: A Linguistic Analysis of the O. J. Simpson Trial. New York: Palgrave Macmillan. Eades, D. (ed.) (1995), Language in Evidence: Issues Confronting Aboriginal and Multicultural Australia. Sydney: UNSW Press. Heffer, C. (2005), The Language of Jury Trial: A Corpus-aided Analysis of Legal-Lay Discourse. Basingstoke and New York: Palgrave Macmillan.

259

HylandK_Chap16_Finals.indd 259

12/1/2010 8:07:30 AM

17

Discourse and Identity Tope Omoniyi

Chapter Overview Introduction Defining ‘Minority’ Identity A Sample Study Key Readings

260 261 263 276

Introduction My task in this chapter is to attempt to synthesize current thinking and research on discourse and identity, as idealistic as that may sound. Language use is a parameter in determining identity, whether of individuals or groups, as research in sociolinguistics, the sociology of language, anthropology, social psychology and other disciplines has shown. In all the work currently being done in identity research we can safely claim as Omoniyi and White (2006) have suggested, that there is general consensus among scholars as to what identity is perceived to be and what its scholarship should cover even if there are variations in the details of process and practice. The work of Le Page and Tabouret-Keller (1985) – Acts of Identity – represents a milestone and a turning point in identity research in language and the communication sciences. We know that human beings as social actors have multiple identities and these are articulated based on salience that is determined by the configuration of social contexts. In other words, one identity may be more appropriate or relevant than others in a given moment in an interaction and rise to the top of a hierarchy of identities (Omoniyi 2006). The hierarchy of identities is a situated social construct that demonstrates the dynamic rather than static property of identification. The norms governing salience operate and are recognized by all members of a network or community. In other words, in the interactional context, movement from one to another stratum on the hierarchy is not arbitrary since such movement is understood to be purposed for a particular objective. 260

HylandK_Chap17_Finals.indd 260

12/2/2010 1:32:11 PM

Discourse and Identity

Its rationale may be unquestionable, but as with all practices and procedures in the empirical sciences, confidence in its deployment will only arise from and grow with use. The argument that we do not cease to be X when salience leads us to claim Y-ness in a given moment in interaction but rather that our X-ness simply moves down the hierarchy of our multiple identities also accounts for the strains of X-ness that may show up in our Y-ness depending on how strongly we feel the need to construct the latter. The stronger the salience, the more prominent the Y-ness that is constructed and less the chances of a strain of X-ness being detected. Nevertheless, the hierarchy remains. Perhaps one of the most fascinating social processes of contemporary times is the structural reconfiguring taking place in vast areas of the world either as a consequence or cause of changing political economies. Brubaker (1996: 148) notes, for instance, on the subject of migration, that ‘massive movements of people have regularly accompanied – as a consequence or as a cause – the expansion, contraction and reconfiguration of political space’. Group identity research is organized around sociological variables such as race, ethnicity, nation, religion, sex and age. It is obvious how these lend themselves to old essentialist approaches to identity study. These were fixed and easily identifiable categories into which individuals and groups were slotted in the process of organizing society. The categories are useful to the extent that they may serve as reference points towards which people may orientate. What contemporary identity research frowns at is taking these reference points as the fixed identities of any person or group. All of the categories are negotiable including biological categories hitherto thought to be fixed, now negotiated through gender reassignment/sex change. Transgender studies have disproved any notions of static identity membership. The nation-state paradigm in conferring citizenship on people creates an idea of identity as fixed and given, once British always British, but the reality is that people experience varying intensity levels in affiliating with the nation. We only realize such intensity discursively. It is apparent in our talk exactly what and how we are feeling about the nation at different points in time granted that in most parts, this may be stable. I want now to turn to the subject of minority identity in Britain and to explore individual and group identity dynamics within that.

Defining ‘Minority’ Identity The term minority identity is very often used loosely to mean any number of persons or groups forming a social unit that is numerically challenged relative to other groups within a polity, or one that has limited access to economic and political power. It is also now used to describe people who subscribe to 261

HylandK_Chap17_Finals.indd 261

12/2/2010 1:32:12 PM

Bloomsbury Companion to Discourse Analysis

supposedly non-mainstream values, thus yielding the distinction gay versus straight for instance. Marshall (1994: 334) says the term ‘has been applied to social groups that are oppressed or stigmatized on the basis of racial, ethnic, biological, or other characteristics’. Bullock et al. (1988: 533) define minorities thus: Groups distinguished by common ties of DESCENT, physical appearance, language, CULTURE or RELIGION, in virtue of which they feel or are regarded as different from the majority of the population in a society. In modern usage the term tends to connote real, threatened or perceived discrimination against minorities, although in exceptional cases (e.g. South Africa) a minority may hold power over a majority. In the 20th century minorities distinguished by other characteristics such as RACE, ethnic identity as immigrants, or sexual preference have joined the range of groups pressing political claims for equality of treatment with that accorded the majority. In this regard perhaps what may be regarded as a classic example of stigmatization and thus an ascription of minority identity is Prince Phillip’s comment that a ramshackle fuse box he spotted on a tour of Racal-MESL, a hi-tech factory near Edinburgh looked ‘as though it was put in by an Indian’ (Daily Star 11 August 1999). The story, which made headline news, was captioned: ‘THE DUKE’S A HAZARD. Phil upsets the Indians’. The paper went on to quote Tricia Marwick (Scottish National Party MSP for Mid-Scotland and Fife) as saying ‘He has a history of similar gaffes and we should not expect him to change now.’ Coincidentally, Jeffery Archer, Conservative Party Member of Parliament, another establishment stalwart had commented to the effect that ‘most black women in Britain used to be fat and badly dressed’ (Daily Star 11 August 1999: p. 2), a statement from which the Conservative Party dissociated itself. Although various other examples of stigmatization and negative stereotyping may be cited, this particular example is most apt in defining minority identity coming in the wake of allegations that institutionalized racism exists in Britain in the aftermath of the Stephen Lawrence Enquiry and the Macpherson Report that followed. An 18-year-old Black youth, Stephen Lawrence had been murdered by a gang of five White youths in 1993. These definitions of minority and the illustrations provided are not exhaustive inasmuch as they do not explicate the positioning of other marginal groups such as environmentalists, gay/lesbian groups, and so on within media discourse. However, the definitions are convenient in mentioning the two most relevant factors to the current discussion – race and religion. I will now examine the media for a localized discussion of the construction and performance of group and individual identities. 262

HylandK_Chap17_Finals.indd 262

12/2/2010 1:32:12 PM

Discourse and Identity

Cesarani and Fulbrook (1996: 5) noted that Britain has a ‘reputedly stable constitutional system, but a weak notion of citizenship and a confused definition of nationality’. It is open to debate whether the weakness and confusion they speak of are a consequence of transforming into a multicultural society. On the other hand, Oommen’s (1997) attempt to precisely define citizenship, nationality and ethnicity in order to ‘avoid ambiguities’ could suggest that the problem is a theoretical one rather than merely a British phenomenon. Following from Cesarani and Fulbrook’s postulate, nationalism would appear to be in conflict with the sentiments that feed minority group identities. The former is rendered seemingly intangible in the face of the intense emotions that sustain its component ‘nationalisms’ – Scottish, Welsh and English, and racialisms – Black, Jewish, Asian and so on (Luthra 1997; Solomos and Back 1995). But whichever way one looks at it, media analysis is useful for exploring the intricacies of inclusion, exclusion, ‘styling the other’ (cf. Rampton 1995) and more generally racism (Cohen and Gardner 1982). Furthermore, studies of representation, social constructions of individual, ethnic and racial identities and their analyses utilize data from media coverage, audience/patronage, discourse style and ideology in meeting their ends (Hall et al. 1978; van Dijk 1991). Riggins’ (1992) five models of ethnic media – integrationist, economic, divisive, pre-emptive and the proselytic – all recognize the mainstream establishment in a sponsor’s role to serve in each case one purpose or another (cited in Gandy 1998: 108). None of these models acknowledge the entrepreneurial capacity of ethnic communities or their recognition of the media as a powerful tool for mobilization and as a representative voice or channel.

A Sample Study The media provides a platform for constructing, contesting, negotiating and conveying plural cultural, religious and linguistic identities. This sample study demonstrates how these identities are constructed and maintained in media discourse with illustrations from a selection of minority newspapers. The newspapers investigated were collected during the spring of 1999 with an update from a smaller sample from early 2010. Arguably, the earlier period was a historical moment for British race politics and ethnic minority identity consciousness. The Macpherson Report on the Stephen Lawrence murder trial had just been released. In the intervening decade, the 9/11 (2001) and 7/7 (2005) terrorist attacks in the United States and Britain respectively had also impacted the notion of minority identity. For instance, race and ethnicity seem to have less significance than religion as an identity variable today and some indication has surfaced on the growing importance of multifaithism. 263

HylandK_Chap17_Finals.indd 263

12/2/2010 1:32:12 PM

Bloomsbury Companion to Discourse Analysis

With reference to the print media, there are newspapers/magazines whose primary ‘constituency’ in terms of coverage of news events is Great Britain (the Guardian, The Times, the Independence). In contrast, there are those that are geographically localized and, therefore, concentrate on events and issues of relevance to the inhabitants of particular regions, towns, cities and borough council areas (the Scotsman, the Reading Chronicle, Ealing Gazette and Newbury Weekly News etc.). There are also those that concentrate on racial/ethnic minority issues. I examined coverage, publishing objectives, language and ideology in the following minority press publications in an attempt to identify how they construct and maintain minority identity in Britain: Amanie, a monthly publication of the Ghana Union London (GUL) and three weeklies, Jewish Chronicle, New Nation and The Voice. I have also taken a cursory look at a 2010 sample of a relatively newer minority publication, the African Voice (London) that did not exist at the time of the initial study and Reading Chronicle (Reading). Overtly or covertly, in the literature on minority identity, a powerful mainstream is often presented as a point of reference against which evaluations of the circumstances of a minority group may be contrasted. The very essence of minority identity derives from the existence of an identifiable and supposedly privileged majority. According to Morris (1982: 79), the purpose of the Black Media Workers Association in Britain included ‘attacking the mainstream white media’ and strengthening the independent Black media ‘because it is one of the small areas of the media industry as a whole over which we exercise some effective control’. The paradigm of analysis employed in racism and anti-racism studies tend to de-emphasize the cooperative and complementary, and foreground ‘difference’ and ‘otherness’ which are the basis for minority identification. The latter establishes a minority reality that is of a different character or texture from mainstream British social, cultural and political reality. As far as the minority press is concerned this notion of otherness may be construed as a form of resistance either to threats of assimilation by the mainstream, or the neglect of issues relating to such groups by the majority media. As Fairclough (1995b: 3) remarked, ‘given the focal position of the mass media in contemporary social systems, there can be little argument about their relevance to the study of sociocultural change’ and indeed the study of non-change in the case of minority cultural maintenance. The treatment of media as symbolic resource whose control is an indicator of the location of power within society (see van Dijk 1991, 1997) makes the majority/minority dichotomy more visible. To argue that minority presses provide a forum for minority expression we must first assume that minority groups either lack access to mainstream media or are negatively represented by it, for which reason a counter-press becomes necessary.

264

HylandK_Chap17_Finals.indd 264

12/2/2010 1:32:12 PM

Discourse and Identity

The discussion will show through an analysis of excerpts from headlines, editorials, feature articles, letters to the editor and advertisements how that counterpositioning is achieved. It is in the process of reacting or resisting that minority identity is firmly asserted and maintained. New Nation, Amanie, The Voice and Jewish Chronicle (JC) are racially defined newspapers from the perspectives of both readership and coverage in Britain. In addition, religion plays a significant role in establishing the identity of JC as a community paper since both race and religion are inseparably webbed in Jewish cultural identity. Under the caption ‘Lacking in Identity’, JC (9 April: p. 6) reported Professor David Newman of Ben-Gurion University as saying that the rejection of Zionism by about 45 per cent of the Israeli population challenges the ideological basis of Israel’s existence. Similarly, Modood (1998) basing on the Fourth National Survey argued that religion is definitely a factor in the new British identity in addition to ethnicity and language especially among minority Asian groups. ‘[E]thnic identification’, Modood says, ‘is no longer necessarily connected to personal participation in distinctive cultural practices, such as those of language, religion or dress’. According to him, there are ‘two principal developments, hybridity and ethno-religious communities’ (1999, personal communication). In relation to the Black and Jewish communities, hybridity may connote a mixture of races, nationalities and even faiths either through adoptive processes or by parentage. The religion factor is the basis for distinguishing between the cultural practices of Sikhs and Hindus for instance. However, religion was already woven into notions of nationality in the original mainstream constructions of Britishness as ‘white, Christian, Anglo-Saxon’ people (Cesarani 1996: 63). The association of the Church of England with the British establishment and of the role of the monarchy as both Head of the Church of England and the United Kingdom parliament are evidence that religion is inherent in mainstream conceptualization of Britishness. Against this background, the minority press constitutes the vehicle for disseminating minority news, discussing minority interests and providing a voice for people who are either seen or see themselves as located on the margins of British society. In this regard, even the names of two of the Black newspapers in this discussion, The Voice and New Nation may be regarded as loaded metaphors. As far as interpretation goes, they imply rights, equality and a need to redefine the British nation. What are the implications of these pursuits for British national, political and/or cultural identity? This and other related questions will steer the direction of the discussion in the remainder of this chapter. There are a number of ways in which minority media construct and maintain the identity of the group they serve or represent. These include delimitation of readership, discourse style, counterpositioning, anchoring to a cultural root

265

HylandK_Chap17_Finals.indd 265

12/2/2010 1:32:12 PM

Bloomsbury Companion to Discourse Analysis

outside Britain, locating, developing and focusing on a notion of ‘home’. I will look at each of these in greater detail.

Delimiting the Readership The idea of a defined and almost exclusive target readership is an extension of the context-determined role of the media. Claims that New Nation is ‘Britain’s No.1 Black Newspaper’ and The Voice is ‘Britain’s Best Black Paper’ indicate the same target readership and in-house generated group identities. This speaks of competition that is being played out within a racial group. Beyond such bold claims in slogans, another marketing ploy, sensationalism and scandalization, is especially associated with the success of tabloid newspapers. However, if a particular scandal or piece of gossip disadvantages a racial group, newspapers serving such communities may capitalize on the occasion to counteract a mainstream-sponsored image of the group. The fact that both New Nation and The Voice newspapers have singled out a racial group for target coverage and readership could arguably be interpreted as a racialization or ethnicization of the media. From the point of view of identity, it can be argued that the referential identity frame is race/ethnicity. In the decade since the data discussed here were collected, the weekly, African Voice, has joined the minority media category and it is described as ‘Britain’s No. 1 African Newspaper’. The significance of this is the fact that it reflects the subdivision of Black Britain into Black British (African), Black British (Caribbean) and Black (Other) in official demographic instruments such as job application forms, health and electoral survey data sheets and so on. The marketing strategy delimits the clientele and clearly identifies a target readership that is racially Black. In Britain, the Black community derives historically from two main sources, the Caribbean island-nations and continental Africa. However, in the context of British national politics, such divisions become irrelevant and the groups are homogenized into a pressure group, which the newspapers seek through their discourses to construct into a Black cultural identity. In the case of JC, ethnic identity comes by way of self-ascription. In its issue of 16 April, under the caption ‘JC Commended’ the paper reported as follows: The Jewish Chronicle was the only ethnic or religious paper which featured in the presentation. Although specialist in its outlook, it comfortably sits at the heart of the Jewish community and is singled out for its high news content and campaigning stance (my emphasis). Judging by the pattern of news coverage in JC, there are three subgroups: British Jewry, Diaspora Jewry and Israelis, with the latter playing anchor to 266

HylandK_Chap17_Finals.indd 266

12/2/2010 1:32:13 PM

Discourse and Identity

the other two. Considering the volume of networking between the three, there does not seem to be a clear-cut division between the three except in situations of conflict. For instance, an Anglo-Jewish organization, the Jewish National Fund (JNF) reviewed its relationship with Keren Kayemet le’Israel Association (KKL) to assert its autonomy and British political statehood (see JC 16 April).

Asserting a Discourse Style With regard to immigrants, the concept of varieties of English permits the isolation and analysis of characteristic forms at phonological, lexical, syntactic, semantic and discourse levels. JC, New Nation, The Voice and Amanie all employ linguistic strategies to define their target readership and by so doing set up a relationship between language and the identity of the communities they serve. One way in which JC demonstrates its commitment to the preservation of Jewish cultural identity is in its use of codeswitching or alternation. Essentially this implies that the readers require bilingual and bicultural competencies in English and Yiddish in order to legitimately assert their membership of the group. I will examine Excerpt 1 from the 30 April edition of JC: Excerpt 1: 1. Denver killer asked the Four Questions at family Seder. (The Four Questions are part of the ritual at the evening meal [seder] on the first night of passover. The youngest male child present traditionally asked them.) 2. A GROUP OF British rabbis has travelled to Stockholm to act as a temporary Beth Din so that 12 Scandinavian couples could receive their long awaited Jewish divorce documents. (Beth Din is a rabbinic court) 3. A TOP Talmudist has challenged the view that Jewish religious law and democracy are fundamentally in conflict. (Talmudist is an expert in talmud, the collection of post-biblical rabbinic interpretations of the biblical laws incumbent on all Orthodox Jews.) 4. It’s a mazal our shop was closed because of Shabbat – Michael Gross shop-owner following the Brick Lane bombing. (mazal = luck; shabbat = sabbath) 5. Jewish communities and Jewish schools across London celebrated Yom Ha’atzmaut last week with activities ranging from Israeli dancing to cookery. (Yom Ha’atzmaut = day of Israel’s independence) 6. Shavuot is said to be judgement day for fruit trees . . . the rest of us have to wait until Yom Kippur! 7. Rabbi Moshe Morgensten, the champion of agunot – ‘chained’ women whose husbands refuse to give them a religious divorce – will be in London, next month. 267

HylandK_Chap17_Finals.indd 267

12/2/2010 1:32:13 PM

Bloomsbury Companion to Discourse Analysis

8. The new pocket prayer book draws attention to the Jewish holidays, Rosh Hashanah (new year) today, and Yom Kippur (day of atonement) on September 20. Full comprehension of the contents of the sentences in 1–5 above requires specific knowledge of aspects of Jewish culture and/or the Yiddish language conveyed by the italicized words. In contrast, sentences 6–8 from the Guardian (10 September, article caption ‘Jews reject Baptist love offensive’) are partially, if not completely, co-textually explained, because the Guardian’s readership includes people to whom these cultural references are inaccessible. As a result, reading and comprehension are not necessarily impeded. In deciding to codeswitch between Jewish and English, publishers of JC have predetermined and defined their readership as an exclusive group. Thus codeswitching, beyond being a sociolinguistic feature of bilingual speech, may be triggered by a more ideological agenda in which case its analysis must take into consideration the social and political identity dynamics of the context. The Voice has a column called ‘Vasta-man Vibration’ which is written in patois (see Excerpt 2, ‘Wedding day blues’ – 29 March) and alludes to the term ‘Rastaman’, a synonym for ‘rastafarian’, a subculture that is symbolic of the West Indian community. Excerpt 2: ‘Wedding day blues’ WHY DID Nubian-minded Monica X choose fi marry ah White man fran Mile End caal Johnny Rotten? Dis ya question bamboozle everybaddi. I ‘n’ I acquainted Monica six years ago and regrettably, mi wedding invitation stipulated dat me tek care ah di Master of Ceremony segment, which was scheduled fi happen during the Anglo-Jamaican supper. Patois is a group-specific culture marker. The notion of hybridity raised by the compound phrase ‘Anglo-Jamaican supper’ is an indication of the existence of a grey area between White Britain and Black Afro-Caribbean. In the context of the wedding ceremony, it takes on a more metaphorical significance. The use of patois is confined to the entertainment section within the paper, which is the arena for popular and global culture. Because the terms are derived from English words, non-speakers/readers of patois may be able to figure out meaning without seeking a translator’s assistance. The fact remains though that the text is not a native English text and remains characteristic of the readership. This kind of group representation is even more complex among the other Black subgroup in Britain, ‘Black Africans’. The diverse language and cultural make-up of Africa’s postcolonial nation-states is reflected in the group. Media 268

HylandK_Chap17_Finals.indd 268

12/2/2010 1:32:13 PM

Discourse and Identity

publications such as West Africa, Ghana Review International, Amanie and so on cater to this diversity and therefore adopt a trans-ethnic perspective. While the nation of origin remains a rallying core, the component nationalisms are equally recognized and advertised through language use. Excerpt 3: From the Poem: ‘Together, we can build a united future’, by A. B. Bodomo Yidandoba, Nananom, Stalwarts I think of you Many markets have come and gone Many moons have come and passed And as the tick tock of Time Tick tocks to Nineteen Ninety-eight I think of you Bayong Hero of the Baobab Naa Gamni Greatest of all architects Osagyefo Kuntunkununku Togbe Adzaladza of Taviafe Nii Nortey Kwakwanya Holders of the fort As the year gets by From far away lands Permit me to send a mesSage to the nation This poem celebrates the linguistic and cultural diversity of Ghana on the occasion of Ghana’s Independence Day anniversary. Its simple message, ‘Happy New Year’, is rendered in several regional languages including Akan, Ewe and Ga. The fact that it is published in Amanie would suggest that its readers, some of whom are British citizens by virtue of their possession of a UK passport, are culturally Ghanaian. This latter identity is the basis of their minority status in Britain and thus the celebration of this reality through poetry may be seen as an enhancement of their difference. Again, as in Excerpt 1, meaningful access is denied to those who lack knowledge of the ‘imported’ linguistic and cultural references. 269

HylandK_Chap17_Finals.indd 269

12/2/2010 1:32:13 PM

Bloomsbury Companion to Discourse Analysis

Counterpositioning In order to assert group identity, reference is sometimes made to the general body of culture that is exclusive to a group, thus counterpositioning it to the Other in the sense of Rattansi’s ‘postmodern frame III’ (1994: 28). A letter published on the Opinion page of JC (23 April), reproduced below as Excerpt 4 illustrates this approach to asserting group membership. It identifies and promotes a set of cultural markers of group difference and identity, and stresses the outsider status of non-members of the group. Excerpt 4: Dangers of Jews dating non-Jews I found Gay Pringle’s article, ‘Guest Appearance’ (JC April 16), distasteful. To interview non-Jews who are the partners of Jewish singles is, effectively, to condone the on-going rise in intermarriage. While you may argue that you are simply reporting a reality, we have to be acutely aware of the dangers of intermarriage. It leads to significant problems within families and, in the long run, is helping to reduce the number of Jews in Britain. We are not necessarily better than others, but we certainly consider ourselves to be special – ‘a kingdom of priests and a light unto the nations.’ (...) I have nothing against non-Jews. They are the majority in this country, and we have to interact with them and befriend them, because friendship creates knowledge, and knowledge promotes harmony. ( . . . ) ABC [Gay Pringle writes: Reporting on a situation is hardly synonymous with condoning it. Indeed, I feel about intermarriage much as Mr. Kreike does. Yet Jews interact with non-Jews, and I believe it is helpful to consider how we are perceived by them. – Ed., JC]. (my emphasis) The rejoinder by Ms Pringle endorsed by the editor of JC may be considered as representing an institutional position on the subject and thus confirms the role of the minority press as protector of group values and culture. The italicized clause adopts the ‘Us/Them’ strategy of counterpositioning the minority to the mainstream or in-group versus out-group memberships. The use of the collective ‘we’ by a compatriot in the utterance ‘We’re not missing much’ addressed to John Diamond, JC columnist and British Jew, as he inspected the prices of gammon and bacon at Safeways on Stamford Hill, London, is often predicated upon an assumption of ‘insider’ versus ‘outsider’ group identities (JC 23 April: p. 29). In a similar vein, the blurb of the film 270

HylandK_Chap17_Finals.indd 270

12/2/2010 1:32:13 PM

Discourse and Identity

Figure 17.1 Front page of The Voice 12–18 April 2010.

‘Solomon and Gaenor’ ran thus: ‘Solomon and Gaenor’ the story of an illicit love affair between a Jewish boy and a Welsh girl’ (JC 30 April: p. 12), condemning as it were interracial relationships to protect or promote cultural purism. This is evident in the choice of ‘illicit’ to describe this particular love affair. We observe the same pattern of pronominalization in stirring up political action as we find on the front page of The Voice (12–18 April 2010): ‘Yes we can make a change – ensure you vote May 6’ (see Figure 17.1 above). 271

HylandK_Chap17_Finals.indd 271

12/2/2010 1:32:14 PM

Bloomsbury Companion to Discourse Analysis

Anchoring The concept of ‘home’ is crucial to a discussion of the spatial location of identity. JC in its ‘Home News’ section carries stories about Jewish events in the United Kingdom thus asserting its identity as an instrument of Anglo-Jewry. The stories take a general perspective on British-Jewry’s relationship with the outside community, other parts of the world, or simply events within it as a subnational group. Complementary to this section is another called Community Chronicle, which focuses on activities in different regional communities such as Brighton, London, Manchester and so on. In contrast, Amanie, in its ‘Home News’ section, carries stories about Ghana and also contains snippets on different UK regions. The issue that needs resolving here is how and the extent to which two immigrant groups in the same country can differ in their notions of ‘home’. In terms of time, Ghanaians have a more recent history of large-scale immigration to Britain. Furthermore some observers will argue that there is a stronger connection between Ghana and Ghanaians than there is between Israel and Diaspora Jews because of the different ports of departure in contemporary times. Most Ghanaian immigrants left Ghana during the crises of the 1970s and 80s to come to the United Kingdom. Anglo-Jewry on the other hand has celebrated the 60th anniversary of the arrival of the first train from Central Europe. Besides, the different nationalities involved make the Jewish community more complex in which case national identity is distinct from cultural identity. With the Ghanaian community, in spite of ethnic differences, on a broad base, cultural identities are less disparate because they evolve from within the same nation. Ironically, JC’s ‘Home News’ pages also carry advertisements relating to Israel. There are two ways of looking at this. First, it may signal an expanded interpretation of ‘home’ to include Israel. Second, British readers who are only interested in this section of the paper come face to face with a subtle ‘think-of-Israel’ propaganda. Take, for example, the following extracts from various editions of JC: World Council of Synagogues helping to find a suitable part-time rabbi from Israel or America’ for the UK’s ninth Masorti synagogue (JC 19 March). Brodetsky nursery introducing Israel into the curriculum. Shalom Orsach, director of overseas development for Melitz (an Israeli agency for non-formal education) agreed to write a regular column for MTP Newsletter to maintain his link with Leeds. Craig Simons from Israel appointed personal assistant to chief executive of the Jewish Welfare Board. ‘Coming form [sic] the outside does give a different perspective than to someone who has lived in Leeds all their life’ in response to a question on what he hoped to bring to his new post (19 March). 272

HylandK_Chap17_Finals.indd 272

12/2/2010 1:32:16 PM

Discourse and Identity

In the last example, Simons constructs himself as an outsider even though his employees may be using him as a channel through which to give their organization some root or anchor. This posture raises the issue of whether there is a greater allegiance between Anglo-Jewry and Anglo-Arab due to their shared British identity than there is between the former and the state of Israel, and between AngloArab and the Arab nations on the other hand. This is a straight contest between ethnicity and nationality. This position can be rationalized from John Diamond’s argument in favour of ‘British Jews’ instead of ‘Jews in Britain’ (JC 23 April 1999). Both New Nation and The Voice are anchored within Britain’s Black communities, personalities and issues, and thus serve to assert a racial identity. In the 10 May (1999) edition of The Voice the ‘Observer’ section carried a story that took on the mainstream press for labelling Reverend Jesse Jackson’s Serbian mission to secure the release of three US airmen held captive in Yugoslavia as a publicity stunt. The paper countered by remarking that Cheri Blair’s (wife of the then British Prime Minister Tony Blair) tears at an Albanian refugee camp was the ultimate publicity stunt since there were people in similar circumstances in Britain who suffered the brunt of Labour’s new immigration and asylum bill. The above excerpts represent what may be described as institutional perspectives on group identity. The Letters section of these newspapers contains readers’ views on issues pertaining to the groups and appears through such perspectives to complement the institutional paradigm of group identity. For instance, the 10 May (1999) edition of The Voice carried a reader’s criticism of Notting Hill the film under the title ‘Ain’t Notting like this’. The writer described the film as ‘make-believe’ because it makes no reference to the Notting Hill Carnival for which the location is renowned. The carnival, the community and the area are central to Black culture and identity in London and as a result, not referencing these facts in the film may be interpreted not only as an act of marginalization, but also as mainstream reconstruction of an established social reality. From a sociolinguistic point of view, the title of the letter sets the identifying stance of the author through its use of syntactic and phonological variants that are obviously Black British vernacular – the replacement of the alveolar fricative /θ/ with the dental stop /t/ in ‘nothing’ spelt ‘notting’. The letter ‘Dangers of dating non-Jews’ (JC 16 April) which I referred to earlier performs a similar function except that it draws solely on cultural references without using overt linguistic characteristics of Anglo-Jewish speech.

Focusing Considering that the objective of minority newspapers is to foreground issues relating to their respective groups, it is understandable that they occupy the headlines and front pages as shown in Table 17.1. 273

HylandK_Chap17_Finals.indd 273

12/2/2010 1:32:17 PM

Bloomsbury Companion to Discourse Analysis

Table 17.1 Distribution of headlines, front-page news and photo images Headlines Paper

Front Page News

Visual Images

In-group

Out-group

In-group

Out-group

In-group

100% 100% 100%

– – –

26 6 7

– – –

98% 92.7% 92.2%

JC NN TV

Out-group 2% 7.3% 7.8%

Table 17.1 shows that in-group news items pertaining to the different client communities dominate the headlines and front pages of the selected newspapers. Similarly, the visual images are predominantly in-group. The portrayal of mainstreamers in a subordinate role is exemplified in this headline news item from New Nation (15 March 1999), which focuses on a Black psychiatrist: IN THE NUT HOUSE! Black shrink treats Stephen murder suspect. One of the five racist thugs suspected of murdering Stephen Lawrence is being treated by a top black psychiatrist, New Nation can reveal. The highly educated Mercedes-driving doctor is a pillar of the medical establishment, the very type Knight and his jobless racist cronies will resent, for both his success and his colour. In these stories, the image of minorities as recipients and mainstreamers as givers of favour is reversed. In this particular instance, the group ascribes some magnanimity to itself in rising above parochialism to embrace and respect professionalism. That edition of New Nation also contained 16 photo frames in ‘The Programme’ section, which deals with sport matters. Except two – Sir Alex Ferguson, and Brazil’s Ronaldo with his girlfriend – all others are of Black athletes from Britain, the United States and Nigeria. Similarly, coverage of national events by minority media is driven by how the community can impact or be affected by the event in question. I shall cite two recent media reports to illustrate this. It was Election Day on 6 May 2010 in the United Kingdom. The Voice (12–18 April 2010) keyed into the ongoing Operation Black Vote Campaign by running on its front page a picture of US President Barack Obama as in Figure 17.1. The image in Figure 17.1 uses intertextuality to invoke two symbolic references. The first is Bob Marley’s popular clarion call to oppressed people to ‘get up, stand up’ for their rights and not give up the fight. The second symbolic reference is to the historic victory achieved by Barack Obama in the 2008 274

HylandK_Chap17_Finals.indd 274

12/2/2010 1:32:17 PM

Discourse and Identity

US presidential election with his slogan incorporated into ‘Yes we can make a change – ensure you vote May 6’ (emphasis mine). Both invoke group history and identity. Similarly, African Voice (14–20 May 2010) ran a front-page story on how the newly formed coalition between the Liberal-Democrats and the Conservatives would impact Minority Britain (see Figure 17.2).

Figure 17.2 Front page of African Voice 14–20 May 2010.

275

HylandK_Chap17_Finals.indd 275

12/2/2010 1:32:18 PM

Bloomsbury Companion to Discourse Analysis

The caption ‘Will Coalition Reflect Multi-Ethnic Britain?’ which has an obvious minority focus appears to be validated by Prime Minister David Cameron and Deputy Prime Minister Nick Clegg photographed waving. That story is complemented by two other stories which link two UK diasporas to their homelands: ‘From a rubbish dump in Lagos to London’s O2’ (p. 11) and ‘10-yr-old survives as Libyan plane crash kills 103’ (p. 17).

Future Directions A number of conclusions can be drawn from the foregoing discussions of media texts vis-à-vis identity research. First, Minority newspapers appear to function as media for the proclamation of minority issues, rights and identities. In order to achieve that end, the communities and publications identify with cultural roots that are external to Britain and from which they generate other identity options. The broadcast media may be more resourceful than the print media as a data source for the analysis of multiple identities and the resulting hierarchy from negotiations within interactions. They have greater scope for accommodating interaction episodes within which identity negotiations take place. The expansion of minority as a category to include new immigrant communities such as the Polish following the reconfiguration of the European Union also calls for a re-examination of some of the issues raised above. The next forage in identity research is beginning and it has to do with the interface between virtual and real social identities. In this regard, Scollon and Scollon’s (2006) nexus analysis paradigm as well as Wenger’s (1998) community of practice before that will be promising frameworks within which to investigate hybridity and interfaces. As globalization continues to ramify, whether cultural or economic, new scholarship will open up on the theme of identity across e-borderlands and new contextualizations of minority.

Key Readings Balibar, E. (2004), We, the People of Europe: Reflections on Transnational Citizenship. Princeton, NJ and Oxford: Princeton University Press. Carmen, L. and Watt, D. (eds) (2010), Language and Identities. Edinburgh: Edinburgh University Press. De Fina, A., Schiffrin, D. and Bamberg, M. (eds) (2006), Discourse and Identity. Cambridge: Cambridge University Press. Omoniyi, T. and White, G. (eds) (2006), Sociolinguistics of Identity. London: Continuum.

276

HylandK_Chap17_Finals.indd 276

12/2/2010 1:32:21 PM

18

Discourse and Race Angel Lin and Ryuko Kubota

Chapter Overview Introduction Discursive Construction of Race: Production of Self and Other Whiteness Studies and its Critiques Elite Discourse and Racism Everyday Discourse and the New Racism Positioning Theory and Storyline Analysis of Racist Online Discourses (Lin and Tong 2009) Conclusion Note Key Readings

277 277 279 280 281 283 289 290 290

Introduction Many researchers have investigated the idea of ‘race’ as a discursive construction, the role of discourse in racialization processes, and the reproduction of racial stereotypes and discrimination of marginalized groups in society. In this body of work, researchers have often drawn on insights from cultural studies, critical theory and postcolonial studies, apart from the analytical tools of different approaches to discourse analysis. In this chapter, we outline key studies which have contributed to this topic. Instead of attempting to be exhaustive, we aim at highlighting intellectual milestones in this area and pointing towards some directions for future research.

Discursive Construction of Race: Production of Self and Other Early critical studies have exposed the historical emergence of the ideology of racism by analysing how the White European colonialist discourse of race 277

HylandK_Chap18_Finals.indd 277

12/1/2010 8:08:37 AM

Bloomsbury Companion to Discourse Analysis

was formed both in conjunction with and in support of the slave trade and slavery institutions. Fryer (1984) in his book on Black history documented and critiqued the discourses of key figures in the colonialist era in the development of the ideology of racism. He pointed out: Once the English slave trade, English sugar-producing plantation slavery, and English manufacturing industry had begun to operate as a trebly profitable interlocking system, the economic basis had been laid for all those ancient scraps of myth and prejudice to be woven into a more or less coherent racist ideology: a mythology of race. (Fryer 1984: 134) Cultural theorists Stuart Hall and Paul Gilroy are two key figures in problematizing the discursive construction of ‘the West and the Rest’ (Hall 1992: 276), and the concept of ‘race’ (Gilroy 2002: 36). Hall pointed out that ‘the West’ is not a fact of geography, but a historical construct discursively produced and reproduced in colonialist discourses and it functions to classify societies and people into different essentialist categories. It produces knowledge about the superior (White) West, and the inferior (non-White) Rest; it discursively constructs both the binary categories of cultural Self and Other and binary sets of knowledge about them (e.g. the civilized, advanced, superior West versus the uncivilized, primitive, inferior Rest). Similarly, ‘race’ is also a concept constructed in these colonialist discourses. As Gilroy (2002) observed: Accepting that skin ‘color’, however, meaningless we know it to be, has a strictly limited basis in biology, opens up the possibility of engaging with theories of signification which can highlight the elasticity and the emptiness of ‘racial’ signifiers as well as the ideological work which has to be done in order to turn them into signifiers in the first place. This perspective underscores the definition of ‘race’ as an open political category, for it is struggle that determines which definition of ‘race’ will prevail and the conditions under which they will endure or wither away. (Gilroy 2002: 36) Edward Said’s Orientalism, a key text in postcolonial theory, shows how the colonialist discourse on the Orient – Orientalism – has constructed a knowledge of the East and power–knowledge relations privileging the West. ‘The Orient was a European invention’, and ‘the Orient has helped to define Europe (or the West) as its contrasting image, idea, personality, experience’ (Said 1985: 1–2). Drawing on Saussure’s linguistic theory (1974) on how signs gain their meanings not through representation of external facts, but through the

278

HylandK_Chap18_Finals.indd 278

12/1/2010 8:08:37 AM

Discourse and Race

setting up of internal contrasts, postcolonial theorists argue that the colonialist has to construct an inferior cultural and racial/ethnic Other in order to know who he/she is (Self): The English are racist not because they hate the Blacks but because they don’t know who they are without the Blacks. They have to know who they are not in order to know who they are. . . . It is a fantastic moment in Fanon’s Black Skin, White Masks when he talks of how the gaze of the Other fixes him in an identity. He knows what it is to be Black when the White child pulls the hand of her mother and says, ‘Look momma, a Black man.’ And he says, ‘I was fixed in that gaze.’ That is the gaze of Otherness. And there is no identity that is without the dialogic relationship to the Other. The Other is not outside, but also inside the Self, the identity. (Hall 1989: 23) Hall (1989) launched a deconstruction of identity and difference in a poststructuralist and post-identity move to do away with essentialist constructions of race and ethnicity and of Self and Other altogether, and to focus on positionality. This has been echoed by recent critiques of Whiteness studies, and arguments for non-essentialist conceptualization of Whiteness in critical education projects (Trainor 2002), as will be discussed in the next section.

Whiteness Studies and its Critiques Whiteness studies were initiated as scholars began to focus on whiteness as discursively and socially constructed and to problematize its link to power and privilege. Works include White Women, Race Matters: The Social Construction of Whiteness by sociologist Ruth Frankenberg (1993), Playing in the Dark: Whiteness and the Literary Imagination by writer and literary critic Toni Morrison (1992), and The Wages of Whiteness by historian David Roediger (1991). However, in recent years, Whiteness studies has been critiqued, chiefly for its ‘difficulty in moving against or away from the master narrative rooted in white supremacy’ (Kubota 2004: 42). Another critique is that there is the danger of essentializing whiteness which is as worrying as the danger of essentializing blackness. As Trainor (2002) points out: [there is] the need to help students articulate antiessentialist identities as whites and to work through the paradoxes of constructing an antiracist white identity. We need to be more aware of the rhetorical frames our pedagogies provide for students as they structure identity . . . Without such

279

HylandK_Chap18_Finals.indd 279

12/1/2010 8:08:37 AM

Bloomsbury Companion to Discourse Analysis

examination, we risk promoting a devastatingly unintended consequence: the development of a conscious, essentialized, and angry white identity predicated on reactionary political values. (Trainor 2002: 647) Essentialist constructions of race/ethnicity, and of Self and Other, however, seem to characterize not only earlier colonialist discourses but also contemporary elite discourses, which constitute our topic in the next section.

Elite Discourse and Racism Key research conducted by T. A. van Dijk uses both linguistic and social psychological approaches to the critical analysis of racist discourses of the White ruling elites in European, British, Australian, New Zealand, North American, Latin American and South African societies (van Dijk 1993, 2005; Wodak and van Dijk 2000). Van Dijk’s classical 1993 study critically analysed racist discourses from all key domains: political discourse, corporate discourse, academic discourse, educational discourse and media discourse. In this important volume, van Dijk both integrated and theoretically elaborated his earlier research on racism and the press (van Dijk 1991) and ethnic prejudice in thought and talk (van Dijk 1987). Van Dijk differentiated between elite racism and popular racism and argued that it is the racist discourses of the elites in different domains of society that provide both the cognitive frameworks and the discursive resources for the reproduction of ethnic stereotypes in everyday talk and thought of the masses. Drawing on Bourdieu (1984, 1988), van Dijk saw these elites as playing an important role in the authorization and legitimation of racist policies and everyday racist practices. Almost outperforming their earlier colonial predecessors, contemporary elites employ a range of sophisticated forms of discourse to legitimate their own social, political, language and economic policies that safeguard their elite status and privilege in society. In line with the thinking of cultural theorists and postcolonial critics who have written earlier on the discursive construction of Self and Other (e.g. Hall 1989; Said 1985), van Dijk, working from his interdisciplinary perspectives informed in particular by linguistic analysis and cognitive schemata theory, saw it as an important empirical project of the discourse analyst to systematically gather the data and analytically demonstrate the discursive and cognitive processes through which racialized Self and Other are constructed both in elite discourses and everyday texts and conversations. While cultural theorists and postcolonial critics are important in providing the insightful observations about the intimate connection between racism and discourse, it is the empirical discourse analyst who systematically gathers and analyses 280

HylandK_Chap18_Finals.indd 280

12/1/2010 8:08:38 AM

Discourse and Race

the discourse data to show the many different ways in which language is recruited and shaped into recurrent, complex formats or patterns that mediate, perpetuate and reproduce racialization and racism in both high and low domains of society. In his methodology, van Dijk drew on different linguistic and research traditions that include: theories of style, rhetoric, narrative, argumentation and conversation, pragmatics, ethnography, and the cognitive and social psychology of text and talk. To illustrate the range of research procedures of such an interdisciplinary approach, van Dijk’s data collection and analysis methods employed in his study on racism and the press might be useful to the reader (see van Dijk 1991: 8–10). In this study, the main data corpus consisted of all types of news discourse that appeared in the British press between 1 August 1985 and 31 January 1986 as well as in the first six months of 1989. This corpus included all news reports, background and feature articles, columns and editorials about ethnic affairs, from both quality press and popular press. A total of 2,700 articles were analysed. All news articles were coded for genre type, size, presence and size of photographs, and overall subject matter (e.g. immigration, race relations, education or crime). Finally, they were coded in terms of a number of propositions formulated in simple clauses (e.g. ‘The Home Secretary said that the riots were criminally inspired’). Then, to study the hypothesized effects of racist reporting on the public, in-depth interviews among 150 newspaper readers (all White people) in Amsterdam and other Dutch cities were also conducted. Van Dijk adopted a sophisticated theoretical framework to inform his analysis which took into account the structures of media discourse, cognitive strategies of news text comprehension and memorization, and the structures and strategies of social representations of the readers.

Everyday Discourse and the New Racism Apart from critical analysis of elite discourse of racism, another important tradition of research in this area is the analysis of (re)production of racial formations and racist ideologies in everyday discourse, in particular, ordinary, mundane conversations and interactions. Without attempting to be exhaustive, three studies will be reviewed here to illustrate the range of theoretical frameworks and methodological tools in this area of study. Whereas earlier colonialist discourses drew on biology to construct racial and ethnic categories, the New Racism (Barker 1981) discourses draw on culture to universalize and essentialize a superior Cultural Self and an inferior Cultural Other. Durrheim and Dixon (2000) analysed how the discourses of White South African holidaymakers justified racial segregation 281

HylandK_Chap18_Finals.indd 281

12/1/2010 8:08:38 AM

Bloomsbury Companion to Discourse Analysis

and criticized social reforms by asserting universal theories of humans as cultural beings to naturalize everyday racist practices. Through a qualitative analysis of the informal interview talk of these holidaymakers, Durrheim and Dixon (2000: 103–4) identified the rhetorical features of lay ontologies of culture. Apart from lay theories about ‘natural’, ‘universal’ human nature, everyday racist discourses are also found to be (re)producing linguistic hierarchies. Anderson (2008) analysed the race talk in her data of ten interviews with women in a large southeast US university town. The interviewees were asked to identify the racial identity of pre-recorded voices. Not only did she find that her interviewees formulated links between speech styles and racial identities, but she also found that they accorded differential values to these speech styles and used this linguistic hierarchy to justify their hegemonic attitudes towards different racial groups (Anderson 2008). The New Racism discourse (Barker 1981) continues to circulate in contemporary society and gets reproduced in everyday interactions. Pagliai’s recent study (2009) drew on Conversation Analysis (CA) (see Kitzinger and Wilkinson this volume) methods in analysing how people engage in repeated conversational agreement to co-construct a racialized image of immigrants in Italy. Pagliai argued that this deployment leads to reinforcement of the racist stances expressed in the conversation itself, and possibly beyond it. The conversationists’ (re)production of ‘the category of the Other’ who are assumed to share the same set of beliefs and actions (e.g. those of a terrorist) is part of an essentializing ‘process of racial formation that can be seen operating at various levels, from everyday conversations, to public political discourse, to the mass media, both in Italy and elsewhere in Europe’, and is part of the New Racism discourses that draw on culture and religion to construct a separate ‘race’ (e.g. ‘Arab’/Muslims) (Pagliai 2009: 568). Pagliai (2009) thus argued that educational programmes created to reduce racist attitudes among the general public should pay attention to not only racist state policies (i.e. elite racism) but also the processes through which racist discourses are (re)produced and circulated in ordinary conversations (i.e. everyday racism). With the global spread of the internet, everyday racist discourses in the new media environment have become a rising concern. In the next section, we shall include excerpts from a study (Lin and Tong 2009) to illustrate how positioning theory and storyline analysis can be used as one of the discourse analysis approaches to analyse how Hong Kong-based TV drama fans co-construct a superior cultural/ethnic Self (Chinese) and an inferior cultural/ethnic Other (Japanese) on an internet fan forum of a Korean TV drama, Dae Jang Geum (the name of a Chosun Dynasty medicine woman). As in the New Racism, culture and ethnicity – two intertwined terms – are proxy for the idea of race (Kubota and Lin 2009). 282

HylandK_Chap18_Finals.indd 282

12/1/2010 8:08:38 AM

Discourse and Race

Positioning Theory and Storyline Analysis of Racist Online Discourses (Lin and Tong 2009) In this study, Lin and Tong (2009)1 draw on positioning theory (Davies and Harré 1990; Harré and Langenhove 1999) to analyse weblog messages to see how different participants use discourse to construct cultural Self and Other. In typical colonial encounters, the colonizer discursively positioned the colonized as a cultural, ethnic and linguistic ‘other’, establishing binary separation of the colonizer and the colonized and asserting the naturalness and primacy of the former (Ashcroft et al. 1998). In both our daily conversations as well as public discourses such discursive construction of Self and Other and of different subject positions for Self and Other routinely occurs. Positioning theory (Davies and Harré 1990) proposes that such subject positions are linked to our discursively constructed storylines which are constantly being negotiated by different parties. For instance, a speaker can position other speakers by adopting a storyline that incorporates a certain interpretation of cultural stereotypes to which other speakers are invited to conform, if they are to continue to interact with the first speaker in a cooperative manner (Davies and Harré 1990). The construction of storyline is central to the establishment and articulation of collective and personal identities, which involves assigning different subject positions (or ‘characters’) to different people in a certain context according to a storyline projected by one’s discourse. By giving others parts in a story, a speaker makes available a subject position that the other speaker normally would take up (Davies and Harré 1990). In projecting storylines, people routinely draw on culturally available stereotypes (or recurring storylines) as resources to position themselves and others. Stereotypes are not pre-existing mental entities or inevitable outcomes of human cognitive functioning; instead they are rhetorical devices that people can use to position themselves and others (Langenhove and Harré 1999). The study focuses on a TV drama fangroup website arising from the Korean TV drama Dae Jang Geum. As positioning involves the process of the ongoing construction of the Self through talk (Davies and Harré 1990; Harré and Langenhove 1999), the discursive practices of these fans show their discursive strategies and tactics in constructing their different/multiple identities with several collective storylines that they co-constructed and sometimes contested.

The Web Discussion Forum of Dae Jang Geum In this study, activities of the Hong Kong Television Broadcasting (TVB)’s web-based discussion forum of Dae Jang Geum were observed regularly from 283

HylandK_Chap18_Finals.indd 283

12/1/2010 8:08:38 AM

Bloomsbury Companion to Discourse Analysis

January 2005 to May 2005. All messages related to cultural and national topics, and construction of identities and storylines are selected for in-depth textual analysis. Although the age of individual members is not specified in their profiles, based on their messages, it is inferred that most of them (approximately 80%) are young students from Hong Kong. The web forum discussion is not confined to Dae or Korean TV dramas. In the following subsections, the textual messages of the members are analysed to understand their discursive acts of constructing Self and Other, and their discursive moves in drawing and shifting boundaries of different subject positions within the ‘storylines’ they created/offered. With some exceptions, almost all of the original messages analysed were written in Chinese characters and they have been translated into English for this study. Viewing rates (number of times the message is read by clicking the topic title) of each message are shown to indicate its popularity.

Re-constructing/re-producing Historical Narratives of a Strong Cultural China as the Centre of Cultural Civilization Dae Jang Geum is set in the political backdrop of Chosun being sandwiched between two strong aggressive powers, China and Japan. These dramatically encoded historical cultural memories and encounters might make popular audience reception difficult in modern-day Mainland China and Japan. However, in the consumption practices of Hong Kong viewers, they seem to exercise selective attention to the dramatic texts and choose to focus on those aspects that they can readily identify with culturally and emotionally. Here is an opening message for the topic: ‘Stella, there are things which you do not know, Chosun is greatly influenced by Chinese culture, Chosun people have been . . . ’ (「Stella，你有所不知，朝鮮深受中國文化影響，朝鮮人一 直 . . . 」). This message shows the writer’s Sino-centric attitudes toward the ‘Great China’ in history. This reproduces and reaffirms the common discourse among Chinese people that China has been historically influencing Korea with its relatively strong and superior culture and loaded with this discourse seems to be the speaker/writer’s sense of cultural pride. When one of the members discovered that the Korean writing used in the dramas looks ‘exactly like the Mandarin characters’, another member called Aviao replied in this way: Yeah, (I) noticed that too. Not really sure but back in those days, China was considered ‘the centre of the world’ so a lot of countries in the vicinity adopted Chinese ways. That’s how Confucianism got spread to Korea and Japan. And I think the writing too. Even the current Japanese writing still incorporate Chinese characters. (Posted on 25 January 2005, viewing rates